Search

Spin-Transfer-Torque (STT) MRAM is configured with high access speed, readily hybrid integration and guaranteed endurance/retention, which becomes one of the most promising candidates for next generation memory. However, the major challenge of STT-MRAM is its high switching energy consumption, accompanied with unreliable writing operation. This work exploits two emerging design paradigms entitled approximation and speculation MRAM. A novel design scheme with timing speculation and approximate storage is proposed to improve energy efficiency in MRAM write operation by reducing write time. Simulation results show that latency and energy loss in entire write operation can be significantly reduced. The proposed technique obtains 84%/57.4% in the symmetric and unidirectional writing operations. Meanwhile, timing speculative method improves the failure/reliability in writing by re-write strategy. Related MRAM macro specification is applied to image processing, which obtains 10% (58.8%-67.2%) of the improvement of precision in image and higher image quality is obtained under the same accuracy rate.

Generative adversarial networks (GANs) are a hot research topic recently. GANs have been widely studied since 2014, and a large number of algorithms have been proposed. However, there is few comprehensive study explaining the connections among different GANs variants, and how they have evolved. In this paper, we attempt to provide a review on various GANs methods from the perspectives of algorithms, theory, and applications. Firstly, the motivations, mathematical representations, and structure of most GANs algorithms are introduced in details. Furthermore, GANs have been combined with other machine learning algorithms for specific applications, such as semi-supervised learning, transfer learning, and reinforcement learning. This paper compares the commonalities and differences of these GANs methods. Secondly, theoretical issues related to GANs are investigated. Thirdly, typical applications of GANs in image processing and computer vision, natural language processing, music, speech and audio, medical field, and data science are illustrated. Finally, the future open research problems for GANs are pointed out.

Automatic image detection is one of the most im- portant areas in computing due to its potential application in numerous real-world scenarios. One important tool to deal with that is called overlap indices. They were introduced as a procedure to provide the maximum lack of knowledge when comparing two fuzzy objects. They have been successfully applied in the following fields: image processing, fuzzy rule-based systems, decision making and computational brain interfaces. This notion of overlap indices is also necessary for applications in which type-2 fuzzy sets are required. In this paper we introduce the notion of type-(2, k) overlap index (k 0, 1, 2) in the setting of type-2 fuzzy sets. We describe both the reasons that have led to this notion and the relationships that naturally arise among the algebraic underlying structures. Finally, we illustrate how type- (2, k) overlap indices can be employed in the setting of fuzzy rule-based systems when the involved objects are type-2 fuzzy sets. This manuscript has been partially supported by Junta de Andalucía by Projects A-FQM-170-UGR20 (Program FEDER Andalucía 2014-2020) and FQM-365 (Andalusian CICYE), and also by Projects PID2020-119478GBI00 and PID2019-108392GB-I00 (AEI/10.13039/501100011033, Ministerio de Ciencia e Innovación), by CNPq (301618/2019-4), FAPERGS (19/2551- 0001660) and grant VEGA 1/0267/21.

An important advantage of optical signal processing is the possibility of performing various operations in parallel. Among the techniques proposed for parallel optical signal processing, frequency division multiplexing (FDM) is a well-known approach, in which different operations are performed in multiple sub-channels associated with different frequency bands. Here, we demonstrate the possibility of performing parallel optical signal processing by frequency division multiplexing of the edge modes of a topological photonic structure. This leads to the realization of a multiple-input-multiple-output (MIMO) topological optical signal processor that is capable of performing several analog functionalities simultaneously, enhancing the speed of computation significantly. Our theoretical and experimental findings open up the venue towards the realization of a new generation of multi-functional topological optical signal processing systems that are ultra-fast, consume little power, and enjoy strong protection against disorder.

Recent transformer-based models, especially patch-based methods, have shown huge potentiality in vision tasks. However, the split fixed-size patches divide the input features into the same size patches, which ignores the fact that vision elements are often various and thus may destroy the semantic information. Also, the vanilla patch-based transformer cannot guarantee the information communication between patches, which will prevent the extraction of attention information with a global view. To circumvent those problems, we propose the Efficient Attention Pyramid Transformer (EAPT) in this paper. More specifically, we first propose Deformable Attention, which learns an offset for each position in patches. Therefore, even with split fixed-size patches, our method can still obtain non-fixed attention information that can cover various vision elements. Then, we design the Encode-Decode Communication module (En-DeC module), which can obtain communication information between all patches to get more complete global attention information. Finally, we also proposed a position encoding specifically for vision transformers, which can be used for patches of any dimensions and any lengths. Extensive experiments on the vision tasks of image classification, object detection, semantic segmentation demonstrate the effectiveness of our proposed model. Furthermore, we also conduct rigorous ablation studies to evaluate the key components of the proposed structure.

High dynamic range (HDR) imaging is a technique that allows an extensive dynamic range of exposures, which is important in image processing, computer graphics, and computer vision. In recent years, there has been a significant advancement in HDR imaging using deep learning (DL). This study conducts a comprehensive and insightful survey and analysis of recent developments in deep HDR imaging methodologies. We hierarchically and structurally group existing deep HDR imaging methods into five categories based on (1) number/domain of input exposures, (2) number of learning tasks, (3) novel sensor data, (4) novel learning strategies, and (5) applications. Importantly, we provide a constructive discussion on each category regarding its potential and challenges. Moreover, we review some crucial aspects of deep HDR imaging, such as datasets and evaluation metrics. Finally, we highlight some open problems and point out future research directions., Comment: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), main and suppl. material

Semantic segmentation is paramount for autonomous vehicles to have a deeper understanding of the surrounding traffic environment and enhance safety. Deep neural networks (DNN) have achieved remarkable performances in semantic segmentation. However, training such a DNN requires a large amount of labelled data at pixel level. In practice, it is a labour-intensive task to manually annotate dense pixel-level labels. To tackle the problem associated with a small amount of labelled data, Deep Domain Adaptation (DDA) methods have recently been developed to examine the use of synthetic driving scenes so as to significantly reduce the manual annotation cost. Despite remarkable advances, these methods unfortunately suffer from the generalisability problem that fails to provide a holistic representation of the mapping from the source image domain to the target image domain. In this paper, we therefore develop a novel ensembled DDA to train models with different up-sampling strategies, discrepancy and segmentation loss functions. The models are, therefore, complementary with each other to achieve better generalisation in the target image domain. Such a design does not only improve the adapted semantic segmentation performance, but also strengthen the model reliability and robustness. Extensive experimental results demonstrate the superiorities of our approach over several state-of-the-art methods.

Tensor fields are useful for modeling the structure of biological tissues. The challenge to measure tensor fields involves acquiring sufficient data of scalar measurements that are physically achievable and reconstructing tensors from as few projections as possible for efficient applications in medical imaging. In this paper, we present a filtered back-projection algorithm for the reconstruction of a symmetric second-rank tensor field from directional X-ray projections about three axes. The tensor field is decomposed into a solenoidal and irrotational component, each of three unknowns. Using the Fourier projection theorem, a filtered back-projection algorithm is derived to reconstruct the solenoidal and irrotational components from projections acquired around three axes. A simple illustrative phantom consisting of two spherical shells and a 3D digital cardiac diffusion image obtained from diffusion tensor MRI of an excised human heart are used to simulate directional X-ray projections. The simulations validate the mathematical derivations and demonstrate reasonable noise properties of the algorithm. The decomposition of the tensor field into solenoidal and irrotational components provides insight into the development of algorithms for reconstructing tensor fields with sufficient samples in terms of the type of directional projections and the necessary orbits for the acquisition of the projections of the tensor field.

Human-machine interactive systems show increasing demand for analysing fast moving objects in high-frame-rate videos. Robust foreground detection, which is able to reduce large amount of redundant background data from high-frame-rate video, becomes the essence to achieve ultra-high-speed human-machine interactions. This paper proposes a local spatial propagation based background model generation, a local linear illumination correction based background model update, and a regional central coordinates and edge keypoints constrained foreground region reselection. The three proposals make up a robust and hardware-friendly foreground detection method. Experimental results prove that the proposed hardware-friendly algorithm achieves high accuracy and robustness on various kinds of challenging cases. Meanwhile, the hardware implementation utilizes little hardware resources and achieves realtime processing of high-frame-rate (784 frame/second) video with the delay less than 1 ms/frame in image processing core. In addition, a practical system is implemented by combing a PC, a high-speed camera and a field programmable gate array (FPGA) for realworld applications. This work will significatively promote the development and application of high-speed human machine interaction. A demo of the proposed vision system working at 784 FPS is available at https://wcms.waseda.jp/em/5f84f75136a6.

In robot visual servoing system, effective detected the region of interest area (ROI) in target images is the first problem should be solved, but the detection is susceptible to an unstructured environment. In this paper, an instance segmentation network algorithm based on Mask Region Convolutional Neural Networks (Mask R-CNN) framework is proposed for ROI image preprocessing. As instance segmentation technology can distinguish complex environmental information, so first with this advantage filter out message such as shape or similar area to target image, then with semantic segmentation technology add special category labels to the filtered images and distinguish different individual instances in similar categories. Finally through the series steps aforementioned, robot vision system can overcome the impact of environmental factors and identify the target image. The proposed method is used to detect target image under five constraints such as occlusion and reflection, result shows the algorithm can effectively deal with challenges that brought by complex constraints, and even can predict the location data of missing information based on some image information. In addition, based on the algorithm proposed in this paper, we used one seven axis robot visual servo platform, executed visual servoing experiment under different unstructured environments, further verifies the effectiveness of our proposed method.

In this article, a periodic self-triggered impulsive (PSTI) control scheme is proposed to achieve synchronization of neural networks (NNs). Two kinds of impulsive gains with constant and random values are considered, and the corresponding synchronization criteria are obtained based on tools from impulsive control, event-driven control theory, and stability analysis. The designed triggering protocol is simpler, easier to implement, and more flexible compared with some previously reported algorithms as the protocol combines the advantages of the periodic sampling and event-driven control. In addition, the chaotic synchronization of NNs via the presented PSTI sampling is further applied to encrypt images. Several examples are also utilized to illustrate the validity of the presented synchronization algorithm of NNs based on PSTI control and its potential applications in image processing.

Data protection is a severe constraint in the heterogeneous IoT era. This article presents a Hardware-Software Co-Simulation of AES-128 bit encryption and decryption for IoT Edge devices using the Xilinx System Generator (XSG). VHDL implementation of AES-128 bit algorithm is done with ECB and CTR mode using loop unrolled and FSM-based architecture. It is found that AES-CTR and FSM architecture performance is better than loop unrolled architecture with lesser power consumption and area. For performing the Hardware-Software Co-Simulation on Zedboard and Kintex-Ultra scale KCU105 Evaluation Platform, Xilinx Vivado 2016.2 and MATLAB 2015b is used. Hardware emulation is done for grey images successfully. To give a practical example of the usage of proposed framework, we have applied it for Biomedical Images (CTScan Image) as a case study. Security analysis in terms of the histogram, correlation, information entropy analysis, and keyspace analysis using exhaustive search and key sensitivity tests is also done to encrypt and decrypt images successfully., Comment: The manuscript is accepted in IEEE Consumer Electronics Magazine . 6 pages in double column format. A total of 4 figures

Frequency response analysis (FRA) suffers from the interpretation of results despite its potential ability to detect faults related to the power transformer windings. This paper presents a technique for interpreting frequency responses, which is based on image processing and deep learning method and called Graph Convolutional Neural Network (CNN). The proposed procedure transfers frequency responses into 2D images through a visualization technique. The gained images are aggregated into a dataset to be used as the CNN input. The proposed technique is applied on frequency responses of two different winding models with Short Circuit (SC) faults. The SC faults with different intensities are applied on different sections of a simulated ladder model winding and a 20kV winding of a 1.6MVA distribution transformer. After determining the frequency response for each faulty case and applying the visualization technique, the precise locating of SC faults is performed by the CNN. Then, the results are analyzed by performance evaluation metrics. At this stage, the high performance of the CNN in the use of 2D images instead of the conventional method is observed. Finally, by testing the high impedance SC faults in different sections of the simulated winding model and applying the suggested method step by step, early detection of the SC fault is also performed in this study. It should be noted that the suggested technique, in addition to its accuracy and high detection speed, can be considered as an important step in automatic interpretation of frequency responses for online monitoring of transformers

To guarantee vehicle platoon driven pattern in heterogeneous nodes of mixed vehicle platoon (composed of connected and automated vehicles and human-driven vehicles, CAVs and HVs) on curved roads, this study develops a combined longitudinal and lateral controller, which comprises of selecting the key points (KPs) from the trajectory points of detected HVs, correcting the reference trajectory and controlling CAVs with the aid of the corrected KPs. To this end, a new concept, called KPs matrix, is proposed to manage the physical components of every KP by using image processing and vehicle-to-infrastructure (V2I) communication technology. Then, a trajectory correction scheme is presented to suppress the influence of nonstandard human-driven behavior by point set mapping approach in Real Variable Function theory. Furthermore, a novel controller is designed by incorporating the corrected KPs matrix and communication time delay. The stability and convergence of the proposed controller are rigorously analyzed based on the Lyapunov-Krasovskii stability theorem. In addition, extensive experiments are conducted to test the performance including three parts: the first part investigates the feasibility of the corrected KPs matrix by analyzing a video on high-way; the next part illustrates the control performance of the proposed controller on handling the cutting-corner issue (i.e. turning in advance), compared with the conventional controller. Meanwhile, the influence of time delay on the control performance is also analyzed in this study. The last implements driver-in-loop comparative experiments such that the performance of the proposed controller on eliminating the influence of nonstandard human-driven behavior is verified.

In image processing, smoke may degrade visibility and deteriorate the performance of high-level vision applications. Therefore, single image smoke removal is crucial for computer vision. Currently, existing smoke removal algorithms mainly leverage handcrafted priors. Moreover, these methods usually apply haze removal methods to perform smoke removal due to the similarity between smoke and haze. However, these methods cannot sufficiently address the degradation of thick smoke and may suffer from residual smoke and color distortion problems due to the non-global and non-homogeneous distribution of smoke. In this paper, to solve the aforementioned problems, an end-to-end deep neural network called DesmokeNet is proposed. We construct a two-stage recovered pipeline to remove the smoke in different thicknesses. The light and thick smoke is first removed locally by the smoke removal network (SRN). The missing pixels in the thick smoke are then recovered by the pixel compensation network (PCN). Moreover, we proposed the thickness-aware pixel loss and the dark channel loss to suppress the residual smoke. To further increase the discriminative ability of the DesmokeNet, we proposed self-attentive feature consensus loss and multi-level contrastive regularization loss to improve the performance of smoke removal. Finally, to train the proposed method, we construct the first large-scale dataset containing synthetic and real-world data. Extensive experiments show that the proposed method outperforms favorably against other state-of-the-art methods quantitatively and qualitatively.

Consistent distribution of shadows on the surface of the shaded PV array enhances the array's harvested capacity. Therefore, in this paper, an Image processing inspired Chaos Map reconfiguration is proposed to mitigate the effect of shading. The proposed reconfiguration is aimed at achieving uniform row current when compared with the other conventional configurations with physical relocation of panels without changing the electrical connections. The efficacy of the proposed method is tested on Short Wide, Short Narrow, Long Wide and Long Narrow shading conditions on the basis of both software and hardware results. The proposed method enhanced output metrics under partially shaded conditions and gives most effective dispersion of shading over entire PV array. This proposed method is also compared with other static existing reconfiguration techniques for the above said shading conditions and the results show tremendous increment in output power under both hardware and software platforms. The results replicate that the proposed configuration can be generalized for any size of an SPV Array and output parameters can be enhanced by physical relocation under any sort of shading conditions.

Respiratory motion is one of the main sources of motion artifacts in positron emission tomography (PET) imaging. The emission image and patient motion can be estimated simultaneously from respiratory gated data through a joint estimation framework. However, conventional motion estimation methods based on registration of a pair of images are sensitive to noise. The goal of this study is to develop a robust joint estimation method that incorporates a deep learning (DL)-based image registration approach for motion estimation. We propose a joint estimation framework by incorporating a learned image registration network into a regularized PET image reconstruction. The joint estimation was formulated as a constrained optimization problem with moving gated images related to a fixed image via the deep neural network. The constrained optimization problem is solved by the alternating direction method of multipliers (ADMM) algorithm. The effectiveness of the algorithm was demonstrated using simulated and real data. We compared the proposed DL-ADMM joint estimation algorithm with a monotonic iterative joint estimation. Motion compensated reconstructions using pre-calculated deformation fields by DL-based (DL-MC recon) and iterative (iterative-MC recon) image registration were also included for comparison. Our simulation study shows that the proposed DL-ADMM joint estimation method reduces bias compared to the ungated image without increasing noise and outperforms the competing methods. In the real data study, our proposed method also generated higher lesion contrast and sharper liver boundaries compared to the ungated image and had lower noise than the reference gated image.

Most compute-intensive applications demand approaches to optimize energy efficiency in CMOS design. Configurable architectures and approximate computing are alternatives to meet the quality and energy requirements in multimedia processing during the runtime. This work proposes a 2D Gaussian filter design to provide multiple quality-power-performance points of operation. The main goal is to design an accuracy-configurable 2D Gaussian filter accelerator at the runtime with a low overhead cost when a new configuration is dynamically selected. The power consumption results show reductions ranging from 9.87% to 58.96%, with an up to 14.8% increase in the hardware area.

A noniterative fractional-order (FO) two-dimensional (2-D) adaptive filtering mechanism is proposed in this article. Quartiles-based adaptive textural segmentation is employed for the calculation of texture-dependent FO. Statistically, the quartiles/quintiles/quantiles are cut points for dividing the range of a probability distribution into continuous intervals with equal probabilities. For the purpose of texture-based isolation of the spatial regions, a 2-D textural map is framed by evaluating spatial entropy by considering a pixel-wise local circular neighborhood. A novel end-to-end framework is proposed for FO texture-dependent image sharpening in an independent manner without influencing the other classes of textural regions. A novel inclusion of texture-wise adaptive gamma correction is also proposed in this article by drafting a mechanism where different kinds of textural regions can be separately processed in an independent manner. A novel quintiles-based multiscale Retinex (MSR) inspired approach for reflectance computation is coined in this article for suppressing environmental artifacts and unbalanced/nonuniform illumination. In this context, various scales required for MSR are themselves computed through quartiles-based intensity levels. The proposed model is highly modular. So, it can also be pipelined in a parallel manner, along with any well-established state-of-the-art contrast enhancement approach. Also, this approach is noniterative and highly robust. It can be proposed as an add-on for several possible (as well as pre-existing) image processing procedures. Rigorous comparative evaluations are performed so that the excellence of the proposed approach can be underlined.

Defect detection of trains is of great significance for operation safety and maintenance efficiency for railway maintenance. Nowadays, China railway system utilizes high-speed line scan cameras to capture images of critical parts of moving trains. The visual inspection on the images still heavily relies on manual interpretation. To reduce the labor requirements, we propose a novel two-stage deep learning based framework for component defect detection of moving trains. The proposed framework is composed of two major successive stages: detecting train components by using our proposed hierarchical object detection scheme (HOD), and detecting component defects based on multiple neural networks and image processing methods. Our proposed HOD can effectively detect and localize train components from large to small in a hierarchical way. Furthermore, a gated feature fusion method that can extract and combine the hierarchical contextual features and spatial contexts is also proposed to improve the performance. To the best of our knowledge, it is the first time in the literature that component defect detection of moving trains is systematically analyzed. Extensive experiments on real images from China railway system have demonstrated that our framework outperforms the state-of-the-art baselines significantly.

Wi-Fi fingerprint-based indoor localization is one of the most practical localization methods, which does not require extra infrastructure and special hardware. However, we need to acquire a dataset with a high-density dataset in the target environment in this framework. To overcome the data acquisition cost problem, we propose a brand new data augmentation for Wi-Fi indoor localization named Between-Location data augmentation (BL data augmentation). We generate the fingerprint data for the whole target environment with high density by only using the sparsely sampled data. Between-Class learning, which is the origin of BL data augmentation and the latest powerful data augmentation method for sound recognition and image processing, mixes two data linearly with normalization; however, this mixing does not make sense in indoor localization because mixed fingerprint has no meaning and the label of indoor localization is not categorical information but physically correlated information. To overcome these two problems, we propose the generative model based on neural networks installed the physical relationship of labels and Wi-Fi fingerprint property. BL data augmentation enables us to reduce data sampled locations while keeping the localization accuracy even if some target locations have no data. From the experimental results, indoor localization methods with BL data augmentation outperform the state-of-the-art data augmentation method on several indoor localization models, whatever the data collection location is dense or sparse. Moreover, the localization with BL data augmentation using 10 % sampled location achieves the same accuracy with localization without data augmentation using all sampled locations.

Traditional fault diagnosis methods are easy to be affected by different working conditions. This paper proposed a motor fault diagnosis method based on visual knowledge, to reduce the impact of changes in working conditions and improve the feature extraction ability. The mapping relationship between actual faults and image intuitive features by symmetrized dot pattern (SDP) and scale invariant feature transform (SIFT) is established in this paper. The fault state is obtained by statistics of the matching point with the dictionary templates generated from signals of normal and unnormal motors. Compared with other machine learning algorithms, this method does not need too much data training and learning. The efficiency of this method is validated by experiments, and the data image processing technology has great industrial application value in the field of motor fault detection or monitoring in the age of intelligence.

Phase-only optical image security technology Based on one-time pad is an encryption and hiding method by introducing random key through phase modulation, which can improve the security of the system to a great extent. The application of traditional phase-only optical image encryption and hiding (POIEH) method is often limited by the amount of data and the quality of decrypted image. In this paper, a single exposure POIEH method using deep learning (DL) is proposed. The original image and corresponding encrypted hidden interferogram acquired with POIEH system are constructed to the train datasets for the learning of an end-to-end designed U-net, then the corresponding relationship between the encrypted hidden interferogram and the reconstructed image is learned and only single-frame encrypted hidden interferogram is needed to perform the reconstruction of POIEH. This method can realize real-time image encryption and hiding and high-quality decryption with only one interferogram. Simulation results and performance analysis show that the proposed method has higher security, stronger generalization ability and robustness.

Overlap functions, as one kind of special binary aggregation functions, have been continuously concerned and studied by many scholars and they have been widely applied in classification, decision making, image processing and fuzzy community detection problems. Meanwhile, migrativity as a vital and particularly interesting property for binary aggregation functions, have been studied in the literature since it was proposed. This paper continues to consider this research topic and mainly focuses on the α-cross-migrativity for overlap functions. Firstly, we introduce the concept of α-cross-migrativity for overlap functions and show some vital properties of α-cross-migrative overlap functions. Secondly, we study the overlap functions that are α-cross-migrative with respect to the minimum overlap function ${O_{M}}$ and obtain an equivalent characterization of them by the ordinal sum of overlap functions. Thirdly, we prove that an overlap function is α-cross-migrative with respect to the p-product overlap function ${O_{p}}$ only if p is equal to 1 and give an equivalent characterization of the overlap functions which satisfy α-cross-migrativity with respect to the 1-product overlap function ${O_{1}}$ also by the ordinal sum of overlap functions. Finally, we extend the α-cross-migrativity of overlap functions to the 0-overlap functions status and get two equivalent characterizations of the 0-overlap functions which are α-cross-migrative with respect to ${O_{M}}$ and an equivalent characterization of the 0-overlap functions which satisfy α-cross-migrativity with respect to ${O_{1}}$ , respectively.

Synthetic aperture radar (SAR), due to its inherent characteristics, will produce speckle noise, which results in the deterioration of image quality, so the removal of speckle in SAR image is very important for the subsequent high-level image processing. In order to balance the relationship between denoising and texture preservation, we propose a multiscale residual dense dual attention network (MRDDANet) for SAR image denoising. This algorithm can effectively suppress the speckle while fully retaining the texture details of the image. In MRDDANet, shallow features are extracted from the noisy images by multiscale modules with different kernel sizes, and then, the extracted shallow features are mapped to the residual dense dual-attention network to obtain the deep features of SAR image. Finally, the final denoising image is generated through global residual learning. MRDDANet has advantages of both multiscale blocks and residual dense dual attention networks. The dense connection can fully extract features in the image, and the dual-channel attention enables MRDDANet to pay more attention to noise information, which is beneficial to remove noise and keep the details of the original image at the same time. Compared with state-of-the-art algorithms, the results of the experiment indicate that our method not only improves various objective indicators but also shows great advantages in visual effects.

Many methods have been developed to improve the performance of image registration. In this letter, we introduce a novel method based on a local affine constraint for remote sensing image registration, which can be widely used in image processing and pattern recognition. Our algorithm has three components. First, we exploit the scale invariant feature transform (SIFT) method to extract feature points and calculate the gradient magnitude to establish feature descriptors with a circular instead of square neighborhood. Second, an initial matching is implemented by the nearest neighbor distance ratio (NNDR) and the fast sample consensus (FSC) algorithm. Finally, fine registration is established using more correct matches obtained by the local affine transformation circular region search algorithm. Experimental results show that the proposed method achieves subpixel accuracy. In addition, both the correct matching rate and registration demonstrate the effectiveness and efficiency of our method.

In seismic exploration methods, imperfect spatial sampling at the surface causes a lack of illumination at the target in the subsurface. The hampered image quality at the target area of interest causes uncertainties in reservoir monitoring and production, which can have a substantial economic impact. Especially in the case of a complex overburden, the impact of surface sampling on target illumination can be significant. The target-oriented acquisition analysis based on wavefield propagation and a known velocity model has been used to provide guidance for optimizing the acquisition parameters. Seismic acquisition design is usually a manual optimization process, with consideration of many aspects. In this study, we develop a methodology that automatically optimizes an irregular receiver geometry when the source geometry is fixed or vice versa. The methodology includes objective functions defined by two criteria: optimizing the image resolution and optimizing the angle-dependent illumination information. We use a two-step parameterization in order to make the problem more linear and, thereby, solve the acquisition design problem by using a gradient descent algorithm. With simple and complex velocity models, we demonstrate that the proposed method is effective, while the involved computational cost is acceptable. Interestingly, the optimization results in our examples show that the conventional uniform geometry already satisfies the resolution requirement, while optimizing for angle coverage can provide a large uplift and is strongly dependent on the velocity model.

Change detection is a key technology in the field of polarimetric synthetic aperture radar (PolSAR) image processing. The current research on the change detection mainly focuses on studying PolSAR images with the same angle or small angle difference, and the angle problem is not considered. However, when the angle difference occurs, especially a large angle difference, some pixels might be falsely detected because the angle difference can affect the polarimetric characteristics. In this letter, we propose a multilevel information fusion-based (MIFB) method, which is suitable for extracting change information from PolSAR images with angle difference. In particular, the proposed method first adopts data resolution correction, then applies an improved feature-based registration algorithm, and finally, incorporates weighted graph theory with the superpixel segmentation algorithm to extract and merge pixel-based and object-based change areas to eliminate false alarms. Experimental results for multitemporal and multiangle PolSAR images reveal that the MIFB method can effectively eliminate false detection caused by angle differences and improve the detection accuracy.

Person Re-IDentification (ReID) aims at re-identifying persons from different viewpoints across multiple cameras. Capturing the fine-grained appearance differences is often the key to accurate person ReID, because many identities can be differentiated only when looking into these fine-grained differences. However, most state-of-the-art person ReID approaches, typically driven by a triplet loss, fail to effectively learn the fine-grained features as they are focused more on differentiating large appearance differences. To address this issue, we introduce a novel pairwise loss function that enables ReID models to learn the fine-grained features by adaptively enforcing an exponential penalization on the images of small differences and a bounded penalization on the images of large differences. The proposed loss is generic and can be used as a plugin to replace the triplet loss to significantly enhance different types of state-of-the-art approaches. Experimental results on four benchmark datasets show that the proposed loss substantially outperforms a number of popular loss functions by large margins; and it also enables significantly improved data efficiency.

Brain diseases are mainly caused by abnormal growth of brain cells that may damage the brain structure, and eventually will lead to malignant brain cancer. An early diagnosis to enable decisive treatment using a Computer-Aided Diagnosis (CAD) system has major challenges, especially accurate detection of different diseases in the magnetic resonance imaging (MRI) images. In this paper, a three-step preprocessing is proposed to enhance the quality of MRI images, along with a new Deep Convolutional Neural Network (DCNN) architecture for effective diagnosis of glioma, meningioma, and pituitary. The architecture uses batch normalization for fast training with a higher learning rate and ease initialization of the layer weights. The proposed architecture is a computationally lightweight model with a small number of convolutional, max-pooling layers and training iterations. A demonstrative comparison between the proposed architecture and other discussed models in this paper is conducted. An outstanding competitive accuracy is achieved of 98.22% overall, 99% in detecting glioma, 99.13% in detecting meningioma, 97.3% in detecting pituitary and 97.14% in detecting normal images when tested on a dataset with 3394 MRI images. Experimental results prove the robustness of the proposed architecture which has increased the detection accuracy of a variety of brain diseases in a short time.

The prime objective of this article is to obtain maximum weld seam positional accuracy and weld quality with varying weld gap values over weld cycles in laser and vision-sensor-assisted robotic arc welding. At first, a vision sensor is used to obtain the seam positions, so that there is no need to teach the robot manually. For a repeatable environment or a similarly designed weld path, the weld seam can then be found by a laser sensor; if there is any error position, it is then notified to a robot controller to move the robot tool center point to the actual position. After that, the output data of the laser sensor in terms of variation of weld gap is used to set the welding parameters at their optimal values by the application of fuzzy logic and NSGA-II algorithm to achieve maximum weld quality. A MOTOMAN MA1440A arc welding robot system is used for experimental investigation of the proposed approach. Furthermore, maximized weld quality in terms of the weld bead size, mechanical, and microstructure properties is achieved by the optimal setting of significant process parameters with respect to the weld gap value obtained as output from the laser sensor. The proposed approach adds more value to the automation process and offline quality improvement of robotic arc welding.

Target detection is always an important topic in the field of hyper/multispectral remote sensing image processing. At present, target detection algorithms in this field are generally limited to processing single-temporal remote sensing data, and they cannot obtain satisfactory results when the spectra of target and background are similar to each other. Recently, a target detection algorithm called filter tensor analysis (FTA), which is specially designed for multitemporal remote sensing data, has been reported and has achieved better detection results in many cases than the traditional single-temporal methods. However, FTA can only extract one target of interest at a time, and it cannot work when there are multiple targets of interest in the image. Therefore, considering that the matrix form of the FTA method is similar to that of the constrained energy minimization (CEM) model, it naturally comes to us that we can combine the tensor filter in FTA and the multiple target constraints to detect multiple targets by fully exploiting the time-series information in multitemporal data. To be specific, through: 1) adding the ``output to one'' constraints to the multiple targets in FTA; 2) applying linear/nonlinear function to the outputs of FTA for the multiple targets; and 3) modifying the autocorrelation matrix in FTA, four multitarget detection algorithms for multitemporal remote sensing data are proposed in this article. Experiments with simulation data and real data both show the effectiveness and superiority of the proposed methods.

The accuracy of the magnetic resonance (MR) image diagnosis depends on the quality of the image, which degrades mainly due to noise and artifacts. The noise is introduced because of erroneous imaging environment or distortion in the transmission system. Therefore, denoising methods play an important role in enhancing the image quality. However, a tradeoff between denoising and preserving the structural details is required. Most of the existing surveys are conducted on a specific MR image modality or on limited denoising schemes. In this context, a comprehensive review on different MR image denoising techniques is inevitable. This survey suggests a new direction in categorizing the MR image denoising techniques. The categorization of the different image models used in medical image processing serves as the basis of our classification. This study includes recent improvements on deep learning-based denoising methods alongwith important traditional MR image denoising methods. The major challenges and their scope of improvement are also discussed. Further, many more evaluation indices are considered for a fair comparison. An elaborate discussion on selecting appropriate method and evaluation metric as per the kind of data is presented. This study may encourage researchers for further work in this domain.

Anomaly detection (AD) is a field of intense research in remote sensing (RS) image processing. Identifying low probability events in RS images is a challenging problem given the high dimensionality of the data, especially when no (or little) information about the anomaly is available a priori. While a plenty of methods are available, the vast majority of them do not scale well to large datasets and require the choice of some (very often critical) hyperparameters. Therefore, unsupervised and computationally efficient detection methods become strictly necessary, especially now with the data deluge problem. In this article, we propose an unsupervised method for detecting anomalies and changes in RS images by means of a multivariate Gaussianization methodology that allows to estimate multivariate densities accurately, a long-standing problem in statistics, and machine learning. The methodology transforms arbitrarily complex multivariate data into a multivariate Gaussian distribution. Since the transformation is differentiable, by applying the change of variables formula, one can estimate the probability at any point of the original domain. The assumption is straightforward: pixels with low estimated probability are considered anomalies. Our method can describe any multivariate distribution, makes an efficient use of memory and computational resources, and is parameter-free. We show the efficiency of the method in experiments involving both AD and change detection (CD) in different RS image sets. For AD, we propose two approaches. The first is using directly the Gaussianization transform and the second is using a hybrid model that combines Gaussianization and the Reed-Xiaoli (RX) method typically used in AD. For CD, we take advantage of the Gaussianization transform and attribute the change to pixels with low probability compared to the first image, instead of those with high difference value typically employed in RS. Results show that our approach outperforms other linear and nonlinear methods in terms of detection power in both anomaly and CD scenarios, showing robustness and scalability to dimensionality and sample sizes.

Compared with the synthetic aperture radar (SAR) image processing theory based on local neighborhood, the nonlocal theory is not limited to a local neighborhood of an image and has great potential in change detection of SAR images. In this study, an approach using ratio-based nonlocal information (RNLI) is proposed for change detection in multitemporal SAR images. First, the RNLI is extracted from a spatial-temporal nonlocal neighborhood where the similarity of two pixels in the nonlocal neighborhood is well characterized by the proposed ratio-based Gaussian kernel function. The parameters of RNLI: noise level and matching window size are adaptively determined to avoid the uncertainty of the change detection result caused by user experience. Second, the difference image is generated by using the RNLI and the ratio operator. Finally, the change map is obtained by segmenting the difference image with a threshold. Experiments conducted on two real datasets and two simulated datasets showed that the proposed method performed better than the other advanced change detection methods, which can better retain the edge information of the changed area while reducing the overall error of the change detection results.

Light detection and ranging (LiDAR) sensors are promising for automated transportation to detect the surrounding environment. However, most LiDAR solutions are complex and bulky. By designing a MEMS-mirror-based LiDAR, we can improve the volume constraints, but MEMS mirrors could limit scanning angles. In this work, we simulated and demonstrated a MEMS LiDAR system to solve the current obstacles. Combining a MEMS mirror and a wide-angle lens into the system, small-volume and large field-of-view (FOV) LiDAR systems can be realized. We use ray tracing optical simulation software to design a pair of aspherical lenses to expand the scanning angle. After the laser beam passes through the wide-angle lens, the FOV can be increased to 104 degrees. The distortion of the wide-angle lens is controlled below 3%, making the scanned image precise to the actual situation. In order to experimentally demonstrate the small-volume MEMS scanning LiDAR, a modular laser rangefinder is used with a MEMS mirror. The entire system of the LiDAR scanner is around 15 cm × 5 cm × 2.5 cm. In the natural light environment for wide-angle LiDAR measurement, the maximum error is less than 2%. Finally, an image processing program is written to convert the scanned data into a 3D point cloud image, and the generated image proves the complete function of the proposed LiDAR.

In the field of remote sensing image processing, clouds heavily affect the quality of the remote sensing images and their application potential. Thus, in recent years, with the prevalence of deep learning techniques used in the field of image processing, many methods have been proposed for cloud removal using single remote sensing images. The existing single-image cloud removal methods suffer from poor generalization capabilities that prevent them from being applied to diverse remote sensing images. Thus, a novel method using a multimodal architecture is proposed which provides multiple most likely outputs for the image and selects the best one through perception-based image quality evaluator (PIQE). In addition, adversarial consistency loss is used to replace cycle consistency loss, which encourages the model to retain more texture information of the original image, and thus the quality of the generated image increases. Experiments demonstrate that the presented method can easily achieve a considerable increase in the peak signal-to-noise ratio and the structural similarity index compared with other methods.

Determining the demographic characteristics of a person post-mortem is a fundamental task for forensic experts, and the dental system is a crucial source of those information. Those characteristics, namely age and sex, can reliably be determined. The mandible and individual teeth survive even the harshest conditions, making them a prime target for forensic analysis. Current methods in forensic odontology rely on time- consuming manual measurements and reference tables, many of which rely on the correct determination of the tooth type. This study thoroughly explores the applicability of deep learning for sex assessment, age estimation, and tooth type determination from x-ray images of individual teeth. A series of models that use state-of-the-art feature extraction architectures and attention have been trained and evaluated. Their hyperparameters have been explored and optimized using a combination of grid and random search, totaling over a thousand experiments and 14076 hours of GPU compute time. Our dataset contains 86495 individual tooth x-ray image samples, with a subset of 7630 images having additional information about tooth alterations. The best-performing models are fine-tuned, the impact of tooth alterations is analyzed, and model performance is compared to current methods in forensic odontology literature. We achieve an accuracy of 76.41% for sex assessment, a median absolute error of 4.94 years for age estimation, and an accuracy of 87.24% to 99.15% for tooth type determination. The constructed models are fully automated and fast, their results are reproducible, and the performance is equal to or better than current state-of-the-art methods in forensic odontology.

Sporadic ocean color events with characteristic spectra, in particular algal blooms, call for quick delivery of high-resolution remote sensing data for further analysis. Motivated by this, we present the mission design for HYPerspectral Smallsat for Ocean observation (HYPSO-1), a 6U CubeSat at 500 km orbital altitude hosting a custom-built pushbroom hyperspectral imager with wavelengths 387-801 nm at 3.33 nm bandpass and a swath width of 70 km. The imager's expected signal-to-noise ratio is characterized for typical open ocean water-leaving radiance which can be flexibly increased by binning pixels. Using geometric principles, the satellite shall execute a slew maneuver during a scan to induce greater overlap in the pixels with a goal to enable better than 100 m spatial resolution. Since high-dimensional hyperspectral data need to be transmitted over limited space-to-ground communications, we have designed a modular FPGA-based onboard image processing architecture that significantly reduces the data size without losing important spatial-spectral information. We justify the concept with a simulated scenario where HYPSO-1 first collects numerous hyperspectral images of a 40 km by 40 km coastal area in Norway and aims to immediately transfer these to nearby ground stations. Using CCSDS123 lossless compression, it takes about one orbital revolution to obtain the complete data product when considering overhead in satellite bus communications and less than 10 min without the overhead. It is shown that even better latency can be achieved with more advanced onboard processing algorithms.

Band selection has been proven to be effective in reducing the dimensionality of the hyperspectral image by finding the most distinctive and informative bands. An essential operation for band selection is to quantify band similarity using metrics, such as entropy and mutual information. For the first time, we proposed that this quantification can also be conducted by borrowing the core ideas from landscape ecology, namely employing landscape metrics. To validate this proposal, we first developed a joint landscape metric and error image approach to quantify the similarity between two bands. Using the quantified similarity and Boltzmann entropy-based information content, we then proposed an efficient priority-based band selection algorithm to search optimal bands. To evaluate the proposed approach, we carried out a comprehensive evaluation involving 80 possible landscape metrics, two methods for quantifying band similarity, four classifiers, and four state-of-the-art, popular approaches as the benchmark. Extensive experimental results demonstrated that the proposed approach exhibited global superiority over these benchmark approaches using all these classifiers. We also found that the best choices of landscape metrics to implement the proposed approach came from the following two categories of metrics: aggregation and diversity. Although this letter presents the first study of its kind in employing landscape metrics for unsupervised band selection, it indicated that landscape metrics might open a door for metric-based approaches for image processing, including band selection.

Image processing via convolutional neural network (CNN) has been developed rapidly for remote sensing technology. Moreover, techniques for accurately extracting building footprints from remote sensed images have attracted considerable interest owing to their wide variety of common applications, including monitoring natural disasters and urban development. Extraction of building footprints can be performed easily by semantic segmentation using U-Net-like CNN architectures. However, obtaining precise boundaries of segmentation masks remains challenging due to various impediments surrounding target objects. In this study, we propose a method to elaborate edges of buildings detected in remote sensed images to enhance the boundaries of segmentation masks. The proposed method adopts holistically nested edge detection (HED), which extracts edge features at an encoder of a given architecture. In the proposed boundary enhancement (BE) module, an extracted edge and segmentation mask are combined, sharing mutual information. To enable the proposed method efficiently to adapt to a wide variety of conditions, we design a distinctive approach adopting a HED unit and BE module, which is applicable to various semantic segmentation networks containing encoder-decoder structures. Experiments were conducted on five different datasets (DeepGlobe, Urban3D, WHU [high-resolution (HR), low-resolution (LR)], and Massachusetts). The results demonstrate that our proposed approaches improved on the performance of prior methods for extracting building footprints. Comparative experiments were conducted on various backbone architectures including U-Net, ResUNet++, TernausNet, and U-shape spatial pyramid pooling (USPP) to ensure the effectiveness of the proposed method. Based on various evaluation metrics and qualitative analysis, our results show that the proposed method achieved improved performance compared with prior methods for all datasets and backbone networks.

A novel spatial singularity-exponent-domain multiresolution imaging (SSMRI) method is proposed in this article. It is aimed at improving the image processing ability for the single-channel synthetic aperture radar (SAR), thereby acquiring the robust image feature for SAR target detection at the very low SNR. Combining the 2-D singularity power spectrum (SPS) and the 2-D pseudo-Wigner-Ville distribution (PWVD), the SSMRI method is derived. Consequently, the single-channel SAR images are transformed to obtain spatial multiresolution SAR images with respect to the singularity exponent. Furthermore, the separability of the SAR image target in the space of the singularity exponent is analyzed, and the separability theorem in the sense of SSMRI is proved. In particular, under the background of GWN and fractal noise, the separability of SAR image targets based on SSMRI processing is studied. In addition, the stability and robustness of SSMRI-SAR feature extraction are demonstrated. On this basis, a maritime SAR ship target detection method based on SSMRI is put forward in the extremely low SNR condition. The experiment results on the SAR-Ship-Dataset indicate that the proposed method is superior in performance to the traditional CFAR or 2-D-SPS method. Especially when SNR =-30 dB, the detection performance with more than 99.4% can be achieved.

In this study, the researchers developed holographic image software for the Polaris, a nongovernmental Taiwanese oceanographic research vessel. It is a survey vessel that was codeveloped through an industry–academia collaboration between National Kaohsiung University of Science and Technology and Dragon Prince Hydro-Survey Enterprise Co. With a weight of 260 tons, length of 36.98 m, and width of 6.80 m, the vessel can travel at a speed of 11 knots. It has undergone underwater rescue and exploration operations and is therefore fairly experienced in such operations. When performing underwater exploration missions, survey vessels are often faced with interferences caused by factors such as current velocity; water temperature, refraction, and spectral conditions; climate; ocean current; presence of algae; and light reflection from schools of fish. Therefore, instantaneous image analysis is imperative for marine exploration. In accordance with the instantaneous recognition needs of the Polaris, the researchers developed artificial-neural-network-based recognition software for rapidly recognizing the category of a detected underwater object. Recognition of shapes in low-resolution underwater images was improved using a neural network resulting in an average recognition rate of 95%. Analysis of variance also indicated that the neural network yielded a significantly higher recognition rate than did manual recognition.

Groundwater is the most significant freshwater source and plays a critical role in the earth's water and energy balance. The lack of groundwater observations with a high spatiotemporal resolution at a global scale hinders our ability to study and model the environment when shallow groundwater has a direct impact on surface soil moisture. This study aims to estimate the spatial and temporal distributions of shallow groundwater-influenced areas at a global scale. We trained an ensemble machine learning algorithm, using outputs from a variably saturated soil moisture flux model, to identify the shallow groundwater occurrence. Model simulations spanned various climate zones and soil types across the globe. The overall accuracy of the algorithm in reproducing the soil moisture flux model results was 95.5%. We applied the algorithm to spaceborne soil moisture observations retrieved by NASA's SMAP satellite and present a global-scale shallow groundwater map derived from the SMAP observations. The derived global distribution of shallow groundwater identifies wetlands, large riparian corridors, and seasonally inundated lowlands. The results showed that 19% of terrestrial land cover had been influenced by shallow groundwater at some point in time during the period of interest (2015–2018). Temporally, shallow groundwater follows an annual cyclic pattern with 2% to 6% of the land surface being influenced globally. This study shows that SMAP observations could be used in estimating shallow groundwater in high spatiotemporal resolution at a global scale, potentially providing invaluable inputs for modeling and environmental monitoring studies.

With the exploration of synthetic aperture radar (SAR), the requirements for its functionality, precision, and response time are inevitably increasing, in which the technical core is the generation, reception, and processing of wideband signals with low time cost. Owing to the excellent performance of modern photonics, such as the ultra-wide bandwidth (UWB), flat response, low transmission loss, fast analog signal processing, and microwave photonics (MWP) promises to be an appropriate solution to improve the capability of SAR in resolution, coverage, and efficiency. However, on the one hand, due to the high sensitivity of optical fiber, moisture, temperature, or physical vibration can produce an unknown extra propagation delay. On the other hand, the wavelength shift effect should be taken into consideration for the UWB system. Thereby, under very high resolution (VHR) circumstance, two-dimensional (2-D) defocus including migration through resolution cells (MTRC) and azimuth phase error (APE) becomes a challenge for MWP SAR imaging. Unfortunately, existing 2-D autofocus approaches concerning motion errors inherently fail for the commonly underlying assumption that the prominent nonsystematic residual range cell migration (RCM) is global in the azimuth time domain. In this article, we analyze the effects of the above-mentioned negative factors in the image processing of range migration algorithm (RMA) and reveal the structural characteristics of the 2-D phase errors. A novel two-step postprocessing compensation strategy is developed, and experiments on real data acquired by an airborne MWP SAR system demonstrate its effectiveness.

Image noise removal is a common problem with many proposed solutions. The current standard is set by learning-based approaches, however these are not appropriate in all scenarios, perhaps due to lack of training data or the need for predictability in novel circumstances. The bitonic filter is a non-learning-based filter for removing noise from signals, with a mathematical morphology (ranking) framework in which the signal is postulated to be locally bitonic (having only one minimum or maximum) over some domain of finite extent. A novel version of this filter is developed in this paper, with a domain that is locally-adaptive to the signal, and other adjustments to allow application to real image sensor noise. These lead to significant improvements in noise reduction performance at no cost to processing times. The new bitonic filter performs better than the block-matching 3D filter for high levels of additive white Gaussian noise. It also surpasses this and other more recent non-learning-based filters for two public data sets containing real image noise at various levels. This is despite an additional adjustment to the block-matching filter, which leads to significantly better performance than has previously been cited on these data sets. The new bitonic filter has a signal-to-noise ratio 2.4dB lower than the best learning-based techniques when they are optimally trained. However, the performance gap is closed completely when these techniques are trained on data sets not directly related to the benchmark data. This demonstrates what can be achieved with a predictable, explainable, entirely local technique, which makes no assumptions of repeating patterns either within an image or across images, and hence creates residual images which are well behaved even in very high noise. Since the filter does not require training, it can still be used in situations where training is either difficult or inappropriate.

Multi-modal retinal image registration plays an important role in the ophthalmological diagnosis process. The conventional methods lack robustness in aligning multi-modal images of various imaging qualities. Deep-learning methods have not been widely developed for this task, especially for the coarse-to-fine registration pipeline. To handle this task, we propose a two-step method based on deep convolutional networks, including a coarse alignment step and a fine alignment step. In the coarse alignment step, a global registration matrix is estimated by three sequentially connected networks for vessel segmentation, feature detection and description, and outlier rejection, respectively. In the fine alignment step, a deformable registration network is set up to find pixel-wise correspondence between a target image and a coarsely aligned image from the previous step to further improve the alignment accuracy. Particularly, an unsupervised learning framework is proposed to handle the difficulties of inconsistent modalities and lack of labeled training data for the fine alignment step. The proposed framework first changes multi-modal images into a same modality through modality transformers, and then adopts photometric consistency loss and smoothness loss to train the deformable registration network. The experimental results show that the proposed method achieves state-of-the-art results in Dice metrics and is more robust in challenging cases.

As all kinds of self-contained display products have become popular, various technologies for improving printing applications have gradually been ignored or even stagnated. However, the number of printed matters in life has not decreased. Every day our lives are full of various printed materials, such as newspapers, magazines, leaflets, product packaging, and so on. The waste and pollution caused by printing technology are usually recessive, and not everyone values it. Unfortunately, people often overlook the waste of resources and environmental pollution caused by the manufacturing process, ink, toner, and printers. The method proposed in this paper uses a color difference model to find a manner that is not easy to notice by the human eye to save ink. The proposed method considers feasibility and practicability, and users can quickly find suitable settings for their applications. Also, through the actual printed pictures, the experimental results show that only 67.57% of the original ink consumption is required to obtain an equivalent print quality that is difficult to perceive by the human eye. Our proposed method will benefit the printing industry and reduce ink consumption on consumer printers, contributing to cost control and being nature-friendly.

In recent years, multimodal biometric systems are increasingly employed in many application field due to several advantages in terms of universality, recognition rate. and security. Among various acquisition technologies, Ultrasound shows important merits, because it allows obtaining volumetric images of the human body and hence a more accurate description of characteristics and to verify liveness. In this work, a multimodal ultrasound recognition system based on the fusion between 3D hand geometry and 3D palmprint features is proposed and experimentally evaluated. The system acquires a volumetric image of the whole hand and for both characteristics, several 2D images are extracted at different depth levels. From each image, 2D features are extracted and then properly combined to achieve a 3D template. Recognition performances are evaluated through verification and identification experiments by employing a homemade database. Experiments are carried out first for the two unimodal biometrics and successively, by fusing the two modalities at score level. Results have shown that fusion is able to dramatically improve the recognition performances of the single biometrics, achieving an Equal Error Rate of 0.08% and an identification rate of 100%.