Search

This paper presents a road detection method for autonomous driving based on an end-to-end neural network model. Our method takes advantage of both the characteristics of road boundary and multi-task learning of a deep convolutional network. By reassigning the label and rebalancing the loss of road pixels, we focus on the learning of hard examples on the boundary to refine its performance. Then, a road geometric transformation-based data augmentation method is proposed to enable the network model to be robust under traffic scenes. Based on these two novel methods, a unified architecture consisting of a shared deep residual encoder network and multi-branch decoder sub-networks is integrated. It adopts road scene classification as a supervised learning task to realize road segmentation and scene classification simultaneously. Experiments demonstrate that the proposed method has achieved the highest MaxF value in most road scenes. Both qualitative and quantitative evaluations on the KITTI-Road benchmark demonstrate our superior performance.

Overhead transmission lines are important lifelines in power systems, and the research and application of their intelligent patrol technology is one of the key technologies for building smart grids. The main reason for the low detection performance of fittings is the wide range of some fittings’ scale and large geometric changes. In this paper, we propose a fittings detection method based on multi-scale geometric transformation and attention-masking mechanism. Firstly, we design a multi-view geometric transformation enhancement strategy, which models geometric transformation as a combination of multiple homomorphic images to obtain image features from multiple views. Then, we introduce an efficient multiscale feature fusion method to improve the detection performance of the model for targets with different scales. Finally, we introduce an attention-masking mechanism to reduce the computational burden of model-learning multiscale features, thereby further improving model performance. In this paper, experiments have been conducted on different datasets, and the experimental results show that the proposed method greatly improves the detection accuracy of transmission line fittings.

Deformable Convolutional Networks (DCNs) can mitigate the inherent limited geometric transformation. We reformulate the spatial-wise attention mechanism using DCNs in this paper for semantic segmentation of high-resolution remote sensing (HRRS) images. It combines the sparse spatial sampling strategy and the long-range relationship modeling capability, namely Deformable Attention Module (DAM). Such locality awareness, more adaptable to HRRS image structures, can capture each pixel’s neighbouring structural information. A reasonable Multi-scale Deformable Attention Net (MDANet) is designed for the HRRS image semantic segmentation with a slightly increased computational cost based on the proposed DAM. Specifically, standard convolutional layers in the raw ResNet50 are equipped with a Deformable Attention Module to control sampling over a broader range of feature levels and aggregate multi-scale context information. The experimental results evaluated on Vaihingen and DeepGlobe Land Cover Classification datasets show that the performance accuracy of MDANet is improved by 7.77% and 8.45% compared with the backbone network (ResNet50) in terms of Miou evaluation, respectively. Furthermore, a DAM can perform better than a global spatial attention mechanism with less computation on the 3×64×64 feature map. Additionally, the added ablation studies demonstrate the effectiveness and efficiency of the deformable attention module and multi-scale strategy, respectively. Moreover, the sensitivity of critical hyperparameters is analyzed.

Passive terahertz (THz) systems can perform the perspective imaging of objects concealed underneath clothing on the human body without a radiometric emitter. In this study, a linear-array passive THz imaging system was designed to realize highly efficient push-broom imaging. However, the acquired THz images are contaminated by severe striping and random noise, which greatly hampers object detection. Therefore, this paper aims to report a dedicated deep learning network for the automatic, accurate, and real-time detection of concealed objects in passive THz images. The main work of this research is threefold. First, a bilateral filter is integrated into a convolutional neural network (CNN) to generate a space-range tunable learning architecture. Second, a space-range grid combined with a geometric transformation matrix is designed to provide a core unit for multi-scale filtering and geometric (MSFG) augmentation. Finally, the one-stage YOLOv5 detector is improved to YOLOv5I by refining the granularity of its prediction units. The proposed approach was tested on onsite passive THz images acquired from various scenes, and the results demonstrate that the proposed MSFG augmentation method can significantly improve the detection accuracy of all the candidate CNNs on the passive THz images. Among them, the YOLOv5I-MSFG was found to exhibit the best performance via a comprehensive assessment of its real-time ability and detection accuracy; the time consumption per frame was found to be ~53 ms on the configured platform, and the accuracies of one-class and two-class detection were found to reach 93.49% and 90.76%, respectively.

Multisource synthetic aperture radar (SAR) image registration is a difficult task in remote sensing due to the influence of speckle noise and geometric distortions between the images. The SAR scale-invariant feature transform (SAR-SIFT) is a prior option to extract features for automatic registration of SAR images. However, most registration methods based on the SAR-SIFT operator only use the image information (the geometric distribution of the pixels making up the feature) and do not consider the influences of the side-looking imagery mode and the topographic relief. In this article, we propose a novel registration method to address the above problems, which combines the optimized SAR-SIFT features and the imaging geometry of the SAR system. The major improvement of the algorithm is the use of the range-Doppler (RD) model in feature matching and geometric transformation estimation. A new local matching method aided by the RD model, transforming the global matching mode to the local one, is first introduced to avoid the gross errors and improve the matching efficiency. Then, an RD-model-based geometric transformation model is presented for multisource SAR images with obvious geometric distortions. In addition, the stationary wavelet transform (SWT) is brought in the feature extraction to optimize the SAR-SIFT keypoints to extract reliable and uniformed features. The experimental results fully demonstrate the applicability of the proposed method for registration of multisource SAR images in different acquiring configurations, especially with different passes.

Identification and detection of traffic identifiers are some of the most challenging tasks in the field of autonomous driving. For object detection, how to obtain deep semantic information and shallow spatial information has always been a permanent subject. This paper proposes a dense network structure for the detection of traffic identifiers; the deep structure of the residual network is used to obtain the high-level semantic information of the detection target; the regression operation is used to obtain the underlying spatial information; the detection structure is higher than the previous The accuracy of the network structure and the generalization ability of the network has been improved; ASPP using the attention mechanism; the main function of ASPP is to obtain a continuous feature map; using the ASPP of the attention mechanism, the prior balance between the receptive field and the feature resolution map is solved; In the paper, a collection of residual networks is used to obtain a variety of acceptance domains; through a regression operation, an effective interactive connection between the receptive field and the acceptance domain is obtained; a deep network using regression residuals allows geometric transformation of the model during data training The ability to improve has been improved; at the same time, the use of a dense network structure has realized the diversity of different receptive fields in the process of multi-scale feature extraction.

A copy-move forgery is a passive tampering wherein one or more regions have been copied and pasted within the same image. Often, geometric transformations, including scale, rotation, and rotation+scale are applied to the forged areas to conceal the counterfeits to the copy-move forgery detection methods. Recently, copy-move forgery detection using image blobs have been used to tackle the limitation of the existing detection methods. However, the main limitation of blobs-based copy-move forgery detection methods is the inability to perform the geometric transformation estimation. To tackle the above-mentioned limitation, this article presents a technique that detects copy-move forgery and estimates the geometric transformation parameters between the authentic region and its duplicate using image blobs and scale-rotation invariant keypoints. The proposed algorithm involves the following steps: image blobs are found in the image being analyzed; scale-rotation invariant features are extracted; the keypoints that are located within the same blob are identified; feature matching is performed between keypoints that are located within different blobs to find similar features; finally, the blobs with matched keypoints are post-processed and a 2D affine transformations is computed to estimate the geometric transformation parameters. Our technique is flexible and can easily take in various scale-rotation invariant keypoints including AKAZE, ORB, BRISK, SURF, and SIFT to enhance the effectiveness. The proposed algorithm is implemented and evaluated on images forged with copy-move regions combined with geometric transformation from standard datasets. The experimental results indicate that the new algorithm is effective for geometric transformation parameters estimation.

This research is a development research (Research & Development) which aims to: determine the feasibility of developing a mathematics module based on Papuan local wisdom in class XI students of SMA Negeri 9 Raja Ampat. The development of this module, the researcher refers to the ADDIE development model (Analysis, Design, Development, Implementation, and Evaluation) with an experimental design using a pre test post test one group design, which uses one research class with a sample of 24 students. The module is developed through 5 stages, namely curriculum and needs analysis, module design, module development, testing the module on a small and wide scale and evaluating the module. Material validation results obtained 92.05% and media validation results with 89.18% meaning that the module is suitable for use in learning by revising. Data analysis results using SPSS 26 for a small scale there are 15 valid questionnaire questions from 20 questions, and the average percentage is 90.21, the reliability value is 0.935, this shows that the questionnaire instrument has a very high level of reliability. The large-scale test for students in class XI is 91.41%, meaning that this module is feasible and effective in improving learning outcomes in geometry transformation material. The n-Gain in the four lesson plans is 0.71, which means that the use of mathematics modules based on Papuan local wisdom is very effective in learning.

This work shows how playing chess creates capacities in the student such as increasing visual memory. This helps to classify information in an orderly manner in the mind and contributes to a better understanding of geometric transformations such as displacements, turns and similarities. This was done with a mixed technique (Quantitative and Qualitative), starting with a structured questionnaire that was applied to 487 students. A case study was carried out with two students (one with and the other without notable chess skills) in two schools in Bogotá-Colombia, with the aim of understanding chess as a tool that can help the teacher to teach mathematics¡. In the quantitative part, data were collected by a structured questionnaire, and in the qualitative part, recordings and transcripts were made of what the two students reported in the case study. So, favorable results were achieved for students who usually play chess, because they show a great capacity for visual memory (in the long and short term) that contributes to a more optimal learning of displacements and similarities in the Cartesian plane. This research shows a powerful tool (chess) that can be used in the teaching of mathematics, thanks to the skills and concepts that are generated in the experience with the game.

Deformable Convolutional Networks (DCNs) are proposed to solve the inherent limited geometric transformation in CNNs, showing outstanding performance on sophisticated computer vision tasks. Though they can rule out irrelevant image content and focus on region of interest to some degree, the adaptive learning of the deformation is still limited. In this paper, we delve it from the aspects of deformable modules and deformable organizations to extend the scope of deformation ability. Concretely, on the one hand, we reformulate the deformable convolution and RoIpooling by reconsidering spatial-wise attention, channel-wise attention and spatial-channel interdependency, to improve the single convolution’s ability to focus on pertinent image contents. On the other hand, an empirical study is conducted on various and general arrangements of deformable convolutions (e.g., connection type) in DCNs. Especially on semantic segmentation, the study yields significant findings for a proper combination of deformable convolutions. To verify the effectiveness and superiority of our proposed deformable modules, we also provide extensive ablation study for them and compare them with other previous versions. With the proposed contribution, our refined Deformable ConvNets achieve state-of-the-art performance on two semantic segmentation benchmarks (PASCAL VOC 2012 and Cityscapes) and an object detection benchmark (MS COCO).

Diabetic eye disease (DED) is a cluster of eye problem that affects diabetic patients. Identifying DED is a crucial activity in retinal fundus images because early diagnosis and treatment can eventually minimize the risk of visual impairment. The retinal fundus image plays a significant role in early DED classification and identification. An accurate diagnostic model’s development using a retinal fundus image depends highly on image quality and quantity. This paper presents a methodical study on the significance of image processing for DED classification. The proposed automated classification framework for DED was achieved in several steps: image quality enhancement, image segmentation (region of interest), image augmentation (geometric transformation), and classification. The optimal results were obtained using traditional image processing methods with a new build convolution neural network (CNN) architecture. The new built CNN combined with the traditional image processing approach presented the best performance with accuracy for DED classification problems. The results of the experiments conducted showed adequate accuracy, specificity, and sensitivity.

BackgroundFunctional imaging especially the SPECT bone scintigraphy has been accepted as the effective clinical tool for diagnosis, treatment, evaluation, and prevention of various diseases including metastasis. However, SPECT imaging is brightly characterized by poor resolution, low signal-to-noise ratio, as well as the high sensitivity and low specificity because of the visually similar characteristics of lesions between diseases on imaging findings.MethodsFocusing on the automated diagnosis of diseases with whole-body SPECT scintigraphic images, in this work, a self-defined convolutional neural network is developed to survey the presence or absence of diseases of concern. The data preprocessing mainly including data augmentation is first conducted to cope with the problem of limited samples of SPECT images by applying the geometric transformation operations and generative adversarial network techniques on the original SPECT imaging data. An end-to-end deep SPECT image classification network named dSPIC is developed to extract the optimal features from images and then to classify these images into classes, including metastasis, arthritis, and normal, where there may be multiple diseases existing in a single image.ResultsA group of real-world data of whole-body SPECT images is used to evaluate the self-defined network, obtaining a best (worst) value of 0.7747 (0.6910), 0.7883 (0.7407), 0.7863 (0.6956), 0.8820 (0.8273) and 0.7860 (0.7230) for accuracy, precision, sensitivity, specificity, and F-1 score, respectively, on the testing samples from the original and augmented datasets.ConclusionsThe prominent classification performance in contrast to other related deep classifiers including the classical AlexNet network demonstrates that the built deep network dSPIC is workable and promising for the multi-disease, multi-lesion classification task of whole-body SPECT bone scintigraphy images.

License plate number detection in drone images is a complex problem because the images are generally captured at oblique angles and pose several challenges like perspective distortion, non-uniform illumination effect, degradations, blur, occlusion, loss of visibility etc. Unlike, most existing methods that focus on images captured by orthogonal direction (head-on), the proposed work focuses on drone text images. Inspired by the Phase Congruency Model (PCM), which is invariant to non-uniform illuminations, contrast variations, geometric transformation and to some extent to distortion, we explore the combination of DCT and PCM (DCT-PCM) for detecting license plate number text in drone images. Motivated by the strong discriminative power of deep learning models, the proposed method exploits fully connected neural networks for eliminating false positives to achieve better detection results. Furthermore, the proposed work constructs working model that fits for real environment. To evaluate the proposed method, we use our own dataset captured by drones and benchmark license plate datasets, namely, Medialab for experimentation. We also demonstrate the effectiveness of the proposed method on benchmark natural scene text detection datasets, namely, SVT, MSRA-TD-500, ICDAR 2017 MLT and Total-Text.

The relationship between spatial conceptions and students' spatial abilities is still rarely studied specifically, even though this is the basis for students to think in learning geometry. This paper aims to explore spatial abilities and the development of spatial ability theory, discusses the relationship between spatial conceptions in students' understanding, and how to develop HLT (Hypothetical Learning Trajectory) in transformation geometry learning . HLT design consists of three stages: initial design, experimental, and retrospective analysis. The results of HLT are then refined into LIT (Local Instructional Trajectory). Then this paper present the empirical results of the perceptions of twenty 9th grade students in one of Islamic private school in Kabupaten Kuningan, West Java, Indonesia, towards the corresponding geometric and math questions . Literature review analysis was used to analyze the retrieved articles. At the end of the paper, we explain and discuss how to apply mathematical conceptions in learning geometry. This research is expected to be a guidance for teachers to develop learning in accordance with the students' spatial thinking process in studying geometry.

3D Speckle tracking techniques are used to quantify cardiac deformation in 3D echocardiographic images. Elastic image registration methods are successful in solving 3D speckle tracking problems. However, a suitable solution should be exploited to deal with the poor spatio-temporal resolution in the echocardiographic images. That is why the registration problem may encounter some challenges in representing accurate features and defining suitable geometric transformation. The strong modeling ability of a fuzzy rule-based inference system can aid the challenge in geometric modeling. This paper, thus, aims to solve the 3D speckle tracking problem in a new scheme through a fuzzy modeling procedure. The algorithm begins to work by extracting a well-suited local feature descriptor, scale- invariant feature transform (SIFT). Then, the relevant features are aligned with sets of fuzzy rules the optimum parameters of which are adaptively learned in the hybrid learning process of adaptive-neuro fuzzy inference system (ANFIS) structure. Applying the adaptive fuzzy method on STRAUS synthetic dataset yields an acceptable tracking error below 1 mm. Further, strain analysis indicates the capacity of the proposed method in discriminating pathological diagnosis from a healthy one.

As an interdisciplinary between quantum computing and image processing, quantum image processing provides more possibilities for image processing due to the powerful parallel computing capabilities of quantum computers. In recent years, quantum image processing attracts more and more researcher’s attention. In order to allow researchers to better understand quantum image processing technology, we have reviewed relevant literature in recent years in the paper. First, the background and mathematical concepts of quantum computing are introduced. Then, the research progress of quantum image processing is sorted out and summarized in the fileds of quantum image representation, geometric transformation, image encryption, edge detection, image segmentation, filtering and compression. Finally, we have discussed the advantages and disadvantages of quantum image processing, and pointed out the potential future research.

The method used in this study was a classroom-based action research that aimed to improve students' mastery of the geometric transformation formulas through the Play Lucky Card technique. This study was conducted in class VIII.2 of SMP Negeri 3 Barru with 23 students. The results achieved after the study was carried out for two cycles were as follows: In the first cycle, there were 40.63% of students who had the ability to master the geometric transformation formulas in the sufficient category. Whereas in the classical results, it was found that the students' average score was 68.06 in the 54.6-64.5 interval or in the sufficient category. In the second cycle, there were 56.25% of students who had the ability to master the mathematical formulas in the high category, 28.13% of students had the ability to master the mathematical formulas in the very high category. From the results of the final evaluation of the second cycle, it was classically found that the students' average score was 76.75 in the 64.6-84.6 interval or in the high category. Based on the results of the final evaluation, the students of class VIII.2 had an absorption amounted to 86.75%, and other data showed that the students of class VIII.2 who scored > 65 (84.38%), were in the high and very high category. Thus, it can be concluded that there was an increase in the ability to master the geometric transformation formulas of class VIII.2 students at SMP Negeri 3 Barru after the implementation of Play Lucky Card technique, there was an increase in motivation and activeness as well as the presence of students following the mathematics learning process. This was reflected in the number (quantity) of students who raised their hands to work on the questions on the blackboard which tended to increase from one meeting to the next. Moreover, almost all students submitted every homework according to predetermined deadlines.

The purpose of this development is to produce an E-module based on Contextual Teaching and Learning in Geometry Transformation material that is valid and practical so that it can be used as one of the mathematics learning resources, especially the material of transformation geometry. The type of research used is research and development (R & D). The model used is the Plomp development model. The stages used in this development model only phase 1 to phase 4, namely the initial investigation phase (preliminary investigation), the design phase (design), the realization/construction phase (realization/construction), the evaluation and revision phase (evaluation and revision). The types of questionnaires used were 2 types, namely an expert validation questionnaire and a student response questionnaire. Based on the results of expert validation the results of the material and the language expert validation percentage were 83.08% (very feasible category), design experts 86.11% (very feasible category) so that the overall material and language validation results and design were 84.56 % (very decent category). The results of the small group trials obtained from the student questionnaire responses to the mathematics E-Module were 88.6% (very practical category). Based on the results of the validity/feasibility and practical group test, it can be concluded that the E-Module mathematics based on Contextual Teaching and Learning (CTL) on the transformation geometry material assisted by Flipbook Maker is very feasible and very practical to be used as one of the sources of mathematics learning, especially on geometry material transformation. Keywords: Development, E-Module, Contextual Teaching and Learning (CTL), Transformation Geometry DOI: http://dx.doi.org/10.23960/mtk/v9i1.pp1-11

In this paper, a new closed-form formulation with universal matrices strongly recommends that Weighted Richardson Extrapolation (WRE) with robust and hourglass controlled one-point quadrature can absolutely replace the conventional Gauss Quadrature in terms of efficiency, accuracy, and speed to find element stiffness matrices of quadrilateral and hexahedral elements. Linearizing the geometric transformation and averaging the material property over an element, using sampling point at the origin (0,0) of a standard mapped 2-square ( ξ , η ) plane for two-dimensional analysis and origin (0,0,0) of a standard mapped 2-cube ( ξ , η , ζ ) system in the hexahedral element (mid-point rule), helps to obtain element stiffness matrix quickly and explicitly through constant universal matrices. This technique is used independently to each of the eight sub-cubes of the mapped 2-cube ( ξ , η , ζ ) of the element for three-dimensional problems and four sub-squares of the mapped 2-square of the element for two-dimensional problems. These matrices are assembled appropriately for a second and better approximation. A weighted addition of the two approximations produces a stiffness matrix as accurate as from conventional Gauss Quadrature. However, finding the stiffness matrix in this way, due to explicit integrations, demands only a third of the time needed for Gauss Quadrature.

Establishing correspondence between two given geometrical graph structures is an important problem in computer vision and pattern recognition. In this paper, we propose a robust graph matching (RGM) model to improve the effectiveness and robustness on the matching graphs with deformations, rotations, outliers, and noise. First, we embed the joint geometric transformation into the graph matching model, which performs unary matching over graph nodes and local structure matching over graph edges simultaneously. Then, the $L_{2,1}$ -norm is used as the similarity metric in the presented RGM to enhance the robustness. Finally, we derive an objective function which can be solved by an effective optimization algorithm, and theoretically prove the convergence of the proposed algorithm. Extensive experiments on various graph matching tasks, such as outliers, rotations, and deformations show that the proposed RGM model achieves competitive performance compared to the existing methods.

Conventionally, geometric computing problems are treated as algebraic computing problems by representing a geometric object in a global reference coordinate system. This approach has two problems. The first problem is that the intuitive view of the interaction of objects is lost, and the second problem is that algebraic computing is prone to errors for degenerated cases related to the interaction of (e.g., notable case of “divided by zero”). In this paper, we propose a new approach to geometric computing especially to analyze the interaction of two objects (e.g., two triangles). The main idea behind this new approach is a specially-tailored local coordinate system for two interacting objects is defined, which makes the projection of the objects on this local coordinate system represent the true geometry of the objects. This idea significantly departs from the conventional approach which is primarily based on the concept of the global coordinate system. Three examples are provided to illustrate the effectiveness of the proposed approach. Among them, one example is related to the robustness of methods for analyzing the relations of two interacting objects, which is still an open issue in the field of geometric computing and suggests that the proposed approach could have some benefit to robustness in analysis.

We propose a new copy-move forgery detection method, which can solve the problems of multiple copy-move forgery, low accuracy and inaccurate tampered region location. First, keypoints and corresponding features of the image are extracted by using AKAZE (accelerated KAZE). Second, features are matched by using the Hamming distance and g2NN which can detect the multiple copy-move forgery. Then, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is used to cluster the keypoints and remove false matching. Finally, PSNR (peak signal-to-noise ratio) and morphological processing is used to locate tampered region accurately. Experimental results show that the proposed method performs well on geometric transformation, post-processing, multiple copy-move forgery, and tampered region localization.

In spite of the good performance of convolutional neural network (CNN) and graph neural network (GNN) in 3D point cloud classification and segmentation at present, to aggregate local information of point clouds and improve the robustness of geometric transformation are still challenging problems. In order to tackle the problems, we propose Geometry Feature Aggregation Network (GFA-Net), which can effectively learn the context information of each point to aggregate local information, so as to enhance the robustness of rotation and translation. Compared with the current popular method GNN that convolves on nearby points in Euclidean space, GFA-Net can better aggregate the geometric features around the points. GFA-Net uses the Laplacian feature mapping to reduce dimensions, and aggregates the nearest neighbor features in the space after dimensionality reduction, and fuses them with the nearest neighbor features of Euclidean space, so as to better obtain the geometric features of each point. Then, points are grouped with geometric features, so that nearby points are insensitive to geometric transformations such as rotation and translation. This method allows GFA-Net to better obtain holistic geometry features, such as symmetry. In addition, we use attention mechanism instead of pooling, so that important neighborhood information can be learned automatically and information loss can be reduced. We conduct extensive experiments on public datasets ModelNet40 and ShapeNet Part. The experimental results show that GFA-Net achieves very good performance, which is very close to the current state-of-the-art methods, and GFA-Net has better robustness.

Geometric and spectral distortions of remote sensing images are key obstacles for deep learning-based supervised classification and retrieval, which are worsened by cross-dataset applications. A learnable geometric transformation model imbedded in a deep learning model has been used as a tool for handling geometric distortions to process close-range images with different view angles. However, a learnable spectral transformation model, which is more noteworthy in remote image processing, has not yet been designed and explored up to now. In this paper, we propose a learnable joint spatial and spectral transformation (JSST) model for remote sensing image retrieval (RSIR), which is composed of three modules: a parameter generation network (PGN); a spatial conversion module; and a spectral conversion module. The PGN adaptively learns the geometric and spectral transformation parameters simultaneously from the different input image content, and these parameters then guide the spatial and spectral conversions to produce a new modified image with geometric and spectral correction. Our learnable JSST is imbedded in the front-end of the deep-learning-based retrieval network. The spatial and spectral-modified inputs provided by the JSST endow the retrieval network with better generalization and adaptation ability for cross-dataset RSIR. Our experiments on four open-source RSIR datasets confirmed that our proposed JSST embedded retrieval network outperformed state-of-the-art approaches comprehensively.

We consider diffusion fields induced by multiple localised and instantaneous sources. We assume a mobile sensor samples the field, uniformly along a piecewise linear trajectory, which is unknown. The problem we address is the estimation of the amplitudes and locations of the diffusion sources, as well as of the trajectory of the sensor. We first propose a method for diffusion source localisation and trajectory mapping (D-SLAM) in 2D, where we assume the activation times of the sources are known and the evolution of the diffusion field over time is negligible. The reconstruction method we propose maps the measurements obtained using the mobile sensor to a sequence of generalised field samples. From these generalised samples, we can then retrieve the locations of the sources as well as the trajectory of the sensor (up to a 2D orthogonal geometric transformation). We then relax these assumptions and show that we can perform D-SLAM also in the case of unknown activation times, from samples of a time-varying field, as well as in 3D spaces. Finally, simulation results on both synthetic and real data further validate the proposed framework.

This paper presents the free vibration and buckling analyses of functionally graded carbon nanotube-reinforced (FG-CNTR) laminated non-rectangular plates, i.e., quadrilateral and skew plates, using a four-nodded straight-sided transformation method. At first, the related equations of motion and buckling of quadrilateral plate have been given, and then, these equations are transformed from the irregular physical domain into a square computational domain using the geometric transformation formulation via discrete singular convolution (DSC). The discretization of these equations is obtained via two-different regularized kernel, i.e., regularized Shannon’s delta (RSD) and Lagrange-delta sequence (LDS) kernels in conjunctions with the discrete singular convolution numerical integration. Convergence and accuracy of the present DSC transformation are verified via existing literature results for different cases. Detailed numerical solutions are performed, and obtained parametric results are presented to show the effects of carbon nanotube (CNT) volume fraction, CNT distribution pattern, geometry of skew and quadrilateral plate, lamination layup, skew and corner angle, thickness-to-length ratio on the vibration, and buckling analyses of FG-CNTR-laminated composite non-rectangular plates with different boundary conditions. Some detailed results related to critical buckling and frequency of FG-CNTR non-rectangular plates have been reported which can serve as benchmark solutions for future investigations.

Cardiovascular diseases (CVDs) are consider the main cause of death today According to World Health Organization (WHO),and because that ECG signal is very important tool in monitoring and diagnosis of these disease , different automatic methods were proposed based on this signal. [1]. The manual analysis of ECG signals is suffered different challenges such as differeculty of detecting and classify waveform of this signal, So, many machine learning methods are explored to describe the anomalies ECG signal accurately . Deep learning (DL) can be used in ECG classification, it can improve the quality of the automatic classification system. In this paper , we have proposed a deep learning classification system by using different layers of convolution, rectifier and pooling operations that can be used to increase feature extraction of ECG signal. We have proposed two models, one is used for input signal of 1-D, in which we designed model for classification csv type of data for ECG signal, while in the second proposed system, we used model for 2-D signal after convert it from its csv type . 2-D signal (ECG image) is used in order to augment the two dimensional signal with different methods to increase the accuracy of the model by training it with geometric transformation of the original input images such as rotation, shearing etc.The results are compared with AlexNet and other models based on the metrics, which are used to measure the performance of the proposed work, the result show that, the proposed models improve the efficiency of the classification in the two systems.

Due to the influence of various conditions and uncertain difficulties for remote sensing images, image registration is still a challenging task. Considering the registration accuracy of pixel level cannot satisfy the requirements of some related applications, we put forward a sub-pixel image registration method based on speeded up robust features and M-estimator sample consensus. It mainly involves four aspects. At first, extract sub-pixel level feature points based on SURF algorithm. Next, obtain the initial matching point pairs based on Sum of Squared Difference and Fast Library for Approximate Nearest Neighbors algorithms. And then, remove the mismatched pair of points based on M-estimator sample consensus algorithm. Finally, calculate geometric transformation matrix based on purified matching points to reach sub-pixel accuracy image registration. Experimental results for several remote sensing image pairs with displacement, noise added, rotation, and different sensors, times and sizes, show that the proposed method can get more anti-interference matches than other methods, and take smaller computational cost in registration process.

In this paper, we present iOrthoPredictor, a novel system to visually predict teeth alignment in photographs. Our system takes a frontal face image of a patient with visible malpositioned teeth along with a corresponding 3D teeth model as input, and generates a facial image with aligned teeth, simulating a real orthodontic treatment effect. The key enabler of our method is an effective disentanglement of an explicit representation of the teeth geometry from the in-mouth appearance, where the accuracy of teeth geometry transformation is ensured by the 3D teeth model while the in-mouth appearance is modeled as a latent variable. The disentanglement enables us to achieve fine-scale geometry control over the alignment while retaining the original teeth appearance attributes and lighting conditions. The whole pipeline consists of three deep neural networks: a U-Net architecture to explicitly extract the 2D teeth silhouette maps representing the teeth geometry in the input photo, a novel multilayer perceptron (MLP) based network to predict the aligned 3D teeth model, and an encoder-decoder based generative model to synthesize the in-mouth appearance conditional on the original teeth appearance and the aligned teeth geometry. Extensive experimental results and a user study demonstrate that iOrthoPredictor is effective in qualitatively predicting teeth alignment, and applicable to the orthodontic industry.

In recent years, watermarking techniques have been under deliberation for their wide variety of applications in various fields. An organization transmits its data through digital communication channels, where the probability of attack is more. To protect the data transmitted through the channel, a novel fuzzy-based digital video watermarking scheme based on Cellular Automata Transform and Singular Value Decomposition (SVD) for copyright protection is proposed. The proposed scheme adopts a mixture of both cellular automata and SVD, which provides copyright protection with high robustness and imperceptibility for multiple transform planes with high processing speed and data redundancy. Moreover, the confidentiality of the proposed scheme is improved with the help of SVD. The watermarked video is experimented with to prove the strength by extracting the watermark image after applying geometric transformation, noise addition, enhancement, and compression attacks. The quantitative performance analysis, which shows improved robustness and imperceptibility of the proposed method compared to the existing methods, is performed by calculating the peak signal-to-noise ratio, structural similarity index, bit error rate, and normalized cross-correlation.

Image enhancement is a technique that manipulates an image to make it more meaningful and effective to user specific problem. In most of the enhancement techniques, input image intensities are transformed into either higher order or lower order intensities according to the designed algorithmic characteristic. But, in certain cases the input intensities might require to be transformed in a balanced combination of both higher and lower order intensity. Moreover, 2D Geometric Transformation is mainly used to transform the objects presents in an image. Here a contemplative fusion of gamma and 2D Geometric Transformation concept has been used for intensity transformation. The proposed method first divides the histogram into three sub-sections according to the homogeneity value representing the dark, gray and bright section of histogram. Then each sub-section is transformed locally using adaptive gamma and 2D Geometric scaling transformation. These transformed sub-sections are merged again by employing 2D translation operation. On the other hand, a global gamma transformation is obtained for entire histogram. At last, the final transformation matrix is obtained by combining previously computed local and global transformation. The comparison of this technique with other state of art technique has been discussed to depict the significance of the proposed method. The proposed method gives a new and innovative dimension of image enhancement.

Safe unmanned ground vehicle navigation in unknown rough terrain is crucial for various tasks such as exploration, search and rescue and agriculture. Offline global planning is often not possible when operating in harsh, unknown environments, and therefore, online local planning must be used. Most online rough terrain local planners require heavy computational resources, used for optimal trajectory searching and estimating vehicle orientation in positions within the range of the sensors. In this work, we present a deep reinforcement learning approach for local planning in unknown rough terrain with zero-range to local-range sensing, achieving superior results compared to potential fields or local motion planning search spaces methods. Our approach includes reward shaping which provides a dense reward signal. We incorporate self-attention modules into our deep reinforcement learning architecture in order to increase the explainability of the learnt policy. The attention modules provide insight regarding the relative importance of sensed inputs during training and planning. We extend and validate our approach in a dynamic simulation, demonstrating successful safe local planning in environments with a continuous terrain and a variety of discrete obstacles. By adding the geometric transformation between two successive timesteps and the corresponding action as inputs, our architecture is able to navigate on surfaces with different levels of friction. Reinforcement learning, autonomous vehicle navigation, motion and path planning.

Locomotion control for quadruped robots is commonly applied on rigid terrains with modelled contact dynamics. However, the robot traversing different terrains is more important for real application. In this paper, a single-leg prototype and a test platform are built. The Cartesian coordinates of the foot-end are obtained through trajectory planning, and then, the virtual polar coordinates in the impedance control are obtained through geometric transformation. The deviation from the planned and actual virtual polar coordinates and the expected force recognized by the ground compliance identification system are sent to the impedance controller for different compliances. At last, several experiments are carried out for evaluating the performance including the ground compliance identification, the foot-end trajectory control, and the comparison between pure position control and impedance control.

Rotary kilns are important industrial plants to process material. In this article, we address the problem of detecting material rings formed in alumina rotary kilns, which leads to high waste and substantial economic loss. This necessitates the use of artificial intelligence to help kiln operators detect material rings promptly. We describe a recently developed system that extracts useful features and detects the presence of material rings efficiently and accurately in real time. Our major contribution is a novel feature extraction method based on geometric transformation and peak localization exploiting the domain knowledge. We also contribute a large data set covering many thousand labeled frames for the evaluation of ring detection. On this data set, we demonstrate that our novel feature extraction method outperforms other alternatives in the literature. Our method is also demonstrated as favorable over deep learning approaches. Our system produces an overall accuracy of nearly 96%, which is acceptable for deployment.

The aim of geometric matching is to extract the geometric transformation parameters between the corresponding images. That is useful for photogrammetric mapping, deformation detection, and flying platform's posture analyses, etc. It is different compare with ordinary feature based image matching succeed by selecting feature points correctly, the proposed method takes all the pixels within the corresponding images to participate the matching procedure for calculating the geometric parameters by least square criterion. The principle of the algorithm, such as the gray corresponding equation, the information quantity inequation and procedure of least square solution are introduced in detail. Particularly, the wavelet analyses for gray signal and calculating the information quantity by signal to noise ratio. Finally, a series of sequential images obtained by a low-altitude helicopter equipped with a video camera was used to test and verify the validity and reliability of the theory and algorithm in this paper. Two typical results are got according to the relative orientation elements model and parallax grid model. The conclusion is got in comparing APM with ordinary feature point method by the information quantity inequation.

Robust estimation, which aims to recover the geometric transformation from outlier contaminated observations, is essential for many remote sensing and photogrammetry applications. This article presents a novel robust geometric model estimation method based on scaled Welsch $q$ -norm ( $l_{q}$ -norm, $0 ). The proposed algorithm integrates a scaled Welsch weight function into the $q$ -norm framework. It, thus, inherits all the advantages of the standard $q$ -norm, i.e., fast and robust. The parameter sensitivity of the standard $q$ -norm is also largely alleviated by integrating such a weight function. These make the proposed algorithm much superior to RANSAC-type methods in real-life applications. We formulate the new cost function as an augmented Lagrangian function (ALF) and divide the ALF into two subproblems [a $q$ -norm penalized least-squares ( $l_{q}$ LS) problem and a weighted least-squares (WLS) problem] by using alternating direction method of multipliers (ADMM) method. For the WLS problem, we introduce a coarse-to-fine strategy into the iterative reweighted least-squares (IRLS) method. We change the weight function by decreasing its scale parameter. This strategy can largely avoid that the solver gets stuck in local minimums. We adapt the proposed cost into classical remote sensing tasks and develop new robust feature matching (RFM), robust exterior orientation (REO), and robust absolute orientation (RAO) algorithms. Both synthetic and real experiments demonstrate that the proposed method significantly outperforms the other compared state-of-the-art methods. Our method is still robust even if the outlier rate is up to 90%.

In recent years, detecting wrecks and submerged sites is a challenging task and time-consuming. A fine technique is required to pick out features from the sea-floor. Side Scan Sonar (SSS) has been used for underwater imaging to detect submerged objects on the sea-floor. Due to the rotational and translational instabilities of SSS tow-fish, the acquired SSS images may be distorted. Hence, this paper concentrates on the geometric transformation analysis for different shadows of SSS images. Based on the intensity-based threshold method, the SSS images are classified as high-shadows, low-shadows, no-shadows, and semi-shadows. Different keypoint detector and mapping algorithms are considered for the extraction of keypoints (features) and matching of those detected points. The point mapping performance was evaluated using Repeatability rate and Mapping rate. It is observed that the Mapping rate of 98.7% is obtained using Scale Invariant Feature Transform (SIFT) with Fast Library for Approximate Nearest Neighbor Search (FLANN) and RANdom Sample Consensus (RANSAC). It is also inferred that the Repeatability rate is 0.8 using Oriented FAST & Rotated BRIEF (ORB) compared with other algorithms considered.

One of the most frequently used types of digital image forgery is copying one area in the image and pasting it into another area of the same image; this is known as the copy-move forgery. To overcome the limitations of the existing Block-based and Keypoint-based copy-move forgery detection methods, in this paper, we present an effective technique for copy-move forgery detection that utilizes the image blobs and keypoints. The proposed method is based on the image blobs and Binary Robust Invariant Scalable Keypoints (BRISK) feature. It involves the following stages: the regions of interest called image blobs and BRISK feature are found in the image being analyzed; BRISK keypoints that are located within the same blob are identified; finally, the matching process is performed between BRISK keypoints that are located in different blobs to find similar keypoints for copy-move regions. The proposed method is implemented and evaluated on the copy-move forgery standard datasets MICC-F8multi, MICC-F220, and CoMoFoD. The experimental results show that the proposed method is effective for geometric transformation, such as scaling and rotation, and shows robustness to post-processing operation, such as noise addition, blurring, and jpeg compression.

In order to improve the contrast of image fusion and highlight the unique characteristics of medical images, a multi-modal medical image fusion algorithm in the framework of non-subsampled contourlet transform (NSCT) is proposed in this paper. Firstly, the computed tomography images and magnetic resonance image are decomposed into low- and high-frequency sub-bands through the NSCT of multi-scale geometric transformation; secondly, for the low-frequency sub-band, the local area standard deviation method is selected or the fusion, while for the high-frequency sub-band, an adaptive pulse coupling neural network model is constructed and the fusion rules are set by the cumulative ignition times of iterative operation in the network; finally, the fusion image is obtained through image reconstruction. Experimental results show that the fusion results of the algorithm in this paper can improve the image fusion quality significantly and it has certain advantages in both visual effects and objective evaluation indexes, which provides a more reliable basis for clinical diagnosis and treatment of diseases.

A new computational procedure for modelling the structural behavior of stiffened plates with symmetry boundary conditions is here presented. It uses two-dimensional finite elements as a way to decrease computational time without losing precision thanks to a relatively small number of elements applied for analyzing out-of-plane displacements (deflections) and stresses. Adding, the constructal design method was included in the procedure, together with the exhaustive search technique, with the scope to optimize the stress/strain status of stiffened plates by design changes. For the purpose, a reference plate without stiffeners was initially design and used as starting point. Part of the volume was reshaped into stiffeners: thickness was reduced maintaining unchanged weight, length and width. The main goal was to minimize strains and stresses by geometric changes. Results demonstrated that, thanks to this design procedure, it is always possible to find an adequate geometry transformation from reference plate into stiffeners, allowing significant improvements in mechanical behavior.

Noninvasive contactless methods for electric power line monitoring based on magnetic field measurement have become an interesting topic for researchers and the electric power industry since introduction of the Smart Grid concept. By measuring and analyzing magnetic field originating from currents in power line conductors, it is possible to detect faults in the network. In medium voltage distribution networks, where a variety of different pole geometries are present, different criteria for fault detection have to be employed for each geometry, which complicates detection and influences accuracy. This paper proposes a novel approach for fault detection in medium voltage distribution networks which is based on processing of signals measured by low cost contactless magnetic field sensors. In order to create a generalized method for fault detection, a sequence of mathematical transformations of the measured magnetic field components is applied. A novel geometric transformation which eliminates influence of pole geometry is introduced, providing signals from which steady-state symmetrical components of the rotating magnetic field are computed. Those components are used as general fault detection criteria. The proposed approach is confirmed to be applicable for different fault types by a set of experiments on three-phase overhead power line model scaled to laboratory conditions.

The image denoising model based on non-local self-similarity prior (NSS) has received extensive attention in recent years because of the repeated structure of natural image patches. Similar patches collected by exploiting NSS prior are sparse, which can be used to estimate potential lowrank subspace. Meanwhile, the modelling of natural images, such as Gaussian mixture models (GMMs), has been successful in all aspects of computer vision by reducing the patterns of image patches. However, the version of its geometric transformation (e.g. rotational transformation) cannot be matched directly by using distance. How to further reduce the patterns of the patches by geometric prior and accelerate rotational matching through parallel calculation is an issue that needs to be solved. In this study, an external guided rotational matching denoising framework is proposed. The proposed framework combines non-local, sparse and low-rank image priors and we design a parallel computing scheme. They demonstrate the performance improvement of the proposed algorithm on images with strong rotational properties and the comparison with traditional state-of-the-art denoising methods. The scalability and effectiveness of the new framework are verified by simulation experiments in public and real datasets.

A geometric transformation method based on discrete singular convolution (DSC) is firstly applied to solve the buckling problem of a functionally graded carbon nanotube (FG-CNT)-reinforced composite skew plate. The straight-sided quadrilateral plate geometry is mapped into a square domain in the computational space using a four-node DSC transformation method. Hence, the related governing equations of plate buckling and boundary conditions of the problem are transformed from the physical domain into a square computational domain by using the geometric transformation-based singular convolution. The discretization process is achieved via the DSC method together with numerical differential and two different regularized kernels such as regularized Shannon’s delta and Lagrange-delta sequence kernels. The accuracy of the present DSC results is first verified, and then, a detailed parametric study is presented to show the impacts of CNT volume fraction, CNT distribution pattern, geometry of the skew plate and skew angle on the axial and biaxial buckling responses of FG-CNTR composite skew plates with different boundary conditions. Some new results related to critical buckling of an FG-CNT-reinforced composite skew plate are also presented, which can serve as benchmark solutions for future investigations.

Photorealistic multi-view face synthesis from a single image is an important but challenging problem. Existing methods mainly learn a texture mapping model from the source face to the target face. However, they fail to consider the internal deformation caused by the change of poses, leading to the unsatisfactory synthesized results for large pose variations. In this paper, we propose a Gated Deformable Face Synthesis Network to model the deformation of faces that aids the synthesis of the target face image. Specifically, we propose a dual network that consists of two modules. The first module estimates the deformation of two views in the form of convolution offsets according to the input and target poses. The second one, on the other hand, leverages the predicted deformation offsets to create the target face image. In this way, pose changes are explicitly modeled in the face generator to cope with geometric transformation, by adaptively focusing on pertinent regions of the source image. To compensate offset estimation errors, we introduce a soft-gating mechanism that enables adaptive fusion between deformable features and primitive features. Extensive experimental results on five widely-used benchmarks show that our approach performs favorably against the state-of-the-arts on multi-view face synthesis, especially for large pose changes.

Existing co-localization techniques significantly lose performance over weakly or fully supervised methods in accuracy and inference time. In this paper, we overcome common drawbacks of co-localization techniques by utilizing self-supervised learning approach. The major technical contributions of the proposed method are two-fold. 1) We devise a new geometric transformation, namely point symmetric transformation and utilize its parameters as an artificial label for self-supervised learning. This new transformation can also play the role of region-drop based regularization. 2) We suggest a heat map extraction method for computing the heat map from the network trained by self-supervision, namely class-agnostic activation mapping. It is done by computing the spatial attention map. Based on extensive evaluations, we observe that the proposed method records new state-of-the-art performance in three fine-grained datasets for unsupervised object localization. Moreover, we show that the idea of the proposed method can be adopted in a modified manner to solve the weakly supervised object localization task. As a result, we outperform the current state-of-the-art technique in weakly supervised object localization by a significant gap.

Semantic segmentation networks usually utilize special pyramid structure after encoder or combine low-level and high-level feature maps in decoder to capture multi-scale context information, which we term them feature combination and feature connection, respectively. However, both such frameworks are less valid with the fixed geometric structure or unsuitable interim. In this paper, we advocate Dynamic Attention Network (DAN) to solve these problems. First, we design a Deformable Attention Pyramid (DAP) module to perform a self-adjustable descriptor of high-level output, which utilizes deformable function to model geometric transformation. With DAP, semantic information can be captured effectively. Second, we propose a Fusing Attention Interim (FAI) module to guide the back-propagation of long-short range information in each level of decoder. We evaluate DAN on the challenging PASCAL VOC 2012 and Cityscapes segmentation benchmarks and find that it achieves state-of-the-art results without post-processing. Our observation can be concluded that the flexible structure that possesses dynamic attention mechanism is beneficial to learn multi-scale context information.

Features are distinctive landmarks of an image. There are various feature detection and description algorithms. Many computer vision algorithms require matching of features from two images. Large number of correct matches with homogeneous distribution in the images is needed for robustness of the image matching. The matches are generally obtained using a feature distance threshold and ambiguous matches are rejected using a ratio test. This paper proposes a method that can be added to image matching pipeline for enhancing homogeneous distribution and increasing the number of matched feature points. After successfully matching an image pair, spatially close feature points go through an elimination process which aims to decrease ambiguity at the second matching step. Then, a coarse geometric transformation between two images is calculated, through which the detected feature points in one image (i.e. the moving image) are projected to the other image (i.e. the fixed image). Then, feature points from the moving image are matched to neighboring feature points of the fixed image within a pre-determined spatial distance. This narrows down the possible candidates and enables less correct matches being rejected because of the ratio test. The effectiveness and feasibility of our method is demonstrated with experiments on images acquired from a drone camera during flight.

Plenty of face detection and recognition methods have been proposed and got excellent results in decades. Common face recognition pipeline consists of: 1) face detection, 2) face alignment, 3) feature extraction, 4) similarity calculation, which are separated and independent from each other. The separated face analyzing stages lead the model redundant calculation and are hard for end-to-end training. In this paper, we proposed a novel end-to-end trainable convolutional network framework for face detection and recognition, in which a geometric transformation matrix was directly learned to align the faces, instead of predicting the facial landmarks. In training stage, our single CNN model is supervised only by face bounding boxes and personal identities, which are publicly available from WIDER FACE [52] dataset and CASIA-WebFace [53] dataset. Tested on Face Detection Dataset and Benchmark (FDDB) [21] dataset and Labeled Face in the Wild (LFW) [19] dataset, we have achieved 89.24% recall for face detection task and 98.63% verification accuracy for face recognition task simultaneously, which are comparable to state-of-the-art results.