Search

• Improved lorenz chaotic systems with higher complexity. • Pixel permutation method based on the image pyramid structure. • Secure and efficient image encryption method. • Galois field multiplication for efficient diffusion of pixels. This paper proposes an improved Lorenz chaotic system and a secure and efficient image encryption method to enhance encryption effectiveness in encrypted images. The proposed improved Lorenz chaotic system addresses the problem of applying the Lorenz chaotic system to image encryption, resulting in weak chaotic characteristics and susceptibility to reconstruction. Dynamic analysis, sensitivity analysis, and randomness testing demonstrate that the improved Lorenz chaotic system exhibits hyperchaotic characteristics, with a maximum Lyapunov exponent of 2.9897. Based on the improved Lorenz chaotic system, this paper proposes an image encryption method that combines image pyramid structure permutation and Galois field diffusion. Unlike most of the current permutation methods limited to a single image layer, this paper proposes a multilayer permutation method based on the image pyramid structure to enhance the permutation effect of image encryption. Although diffusion based on Galois field multiplication operation is efficient and secure, it is less effective in encrypting pixel points with a pixel value of '0′. To address this issue, this paper incorporates DNA computing into diffusion based on Galois field operations, enabling even pure black images to achieve better encryption effectiveness. Experimental results demonstrate that the encryption method proposed in this paper effectively conceals information contained in the plain image. The global Shannon entropy of the encrypted Lena image can reach 7.9975, indicating a high level of randomness and complexity. Notably, even a slight alteration, such as changing a single pixel, results in a significant divergence, with 99.6307 % of the cipher image's pixels being distinct. Moreover, it effectively withstands analysis from various attacks. Therefore, the encryption method proposed in this paper can be effectively applied to grayscale image encryption scenarios requiring relatively high security and encryption efficiency, such as remote sensing image encryption and personal privacy image encryption. [ABSTRACT FROM AUTHOR]

Copyright of Computer Measurement & Control is the property of Magazine Agency of Computer Measurement & Control and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Deep learning-based medical image segmentation requires a large number of labeled data to train the model. Obtaining large-scale labeled medical image datasets is time-consuming and expensive. In contrast, it is easy to obtain unlabeled data, which also deserve to be effectively explored to improve the segmentation quality. To solve this problem, we proposed a semi-supervised deep learning method based on Generative Adversarial Network (GAN) in combination with a pyramid attention mechanism and transfer learning (TP-GAN). In this work, TP-GAN consisted of a generator (segmentation network) and a discriminator (evaluation network). The generator adopted the encoder-decoder architecture for image segmentation (the output was called the predicted map), and the discriminator adopted convolutional neural network (CNN) to evaluate the quality of the predicted map. Through adversarial training between generator and discriminator, TP-GAN could achieve high segmentation quality since discriminator guides the generator to generate more accurate segmentation maps with more similar distribution as ground truth for unlabeled data in semi-supervised learning. Furthermore, the encoder in generator utilized the VGG16 model which had been trained for image classification on ImageNet data, meanwhile constituted a new segmentation model with the decoder. Transfer learning strategy could reduce the training time and overcome the limitation of small-scale labeled data in semi-supervised learning. And the generator used image pyramid attention mechanism to extract more detailed features to enhance the information of feature maps. The proposed TP-GAN model and other segmentation models were trained and tested on two different datasets (Hippocampus and Spleen). The results demonstrated that TP-GAN could achieve higher segmentation accuracy on the Hippocampus and Spleen than other semi-supervised segmentation methods based on different evaluation metrics (Dice, IoU, HD, and RVE). The proposed TP-GAN model could effectively utilize the unlabeled data to improve the segmentation quality. And TP-GAN could relieve the burden of a tedious image annotation process and reduce the influence of physicians' subjective experiences in clinical practice. [ABSTRACT FROM AUTHOR]

Citrus is an important commercial crop in many areas. The management and planning of citrus growing can be supported by timely and efficient monitoring of citrus-growing regions. Their complex planting structure and the weather are likely to cause problems for extracting citrus-growing regions from remote sensing images. To accurately extract citrus-growing regions, deep learning is employed, because it has a strong feature representation ability and can obtain rich semantic information. A novel model for extracting citrus-growing regions by UNet that incorporates an image pyramid structure is proposed on the basis of the Sentinel-2 satellite imagery. A pyramid-structured encoder, a decoder, and multiscale skip connections are the three main components of the model. Additionally, atrous spatial pyramid pooling is used to prevent information loss and improve the ability to learn spatial features. The experimental results show that the proposed model has the best performance, with the precision, the intersection over union, the recall, and the F1-score reaching 88.96%, 73.22%, 80.55%, and 84.54%, respectively. The extracted citrus-growing regions have regular boundaries and complete parcels. Furthermore, the proposed model has greater overall accuracy, kappa, producer accuracy, and user accuracy than the object-oriented random forest algorithm that is widely applied in various fields. Overall, the proposed method shows a better generalization ability, higher robustness, greater accuracy, and less fragmented extraction results. This research can support the rapid and accurate mapping of large-scale citrus-growing regions. [ABSTRACT FROM AUTHOR]

In visual SLAM, the extraction of feature points is an important factor that affects global real-time localization and mapping efficiency. This paper studies visual ORB-SLAM2 algorithm and proposes an adaptive meshing method to optimize the efficiency of feature point extraction. Meshing the image pyramid layer improves the speed of feature point extraction. This study conducts monocular (MONO) and RGB-D tests on the TUM dataset. The results show the average feature point extraction time per frame increases by 8%-10% while the absolute trajectory error is down by over 5% respectively. Meanwhile, the RGB-D dense point cloud construction thread is added to the adaptive grid algorithm, and the outlier removal filter and voxel grid filter are employed to reduce the point cloud scale and realize dense mapping. The experimental results on the TUM dataset show the method is highly effective in indoor dense mapping. [ABSTRACT FROM AUTHOR]

Medical image fusion aims to integrate complementary information from multimodal medical images and has been widely applied in the field of medicine, such as clinical diagnosis, pathology analysis, and healing examinations. For the fusion task, feature extraction is a crucial step. To obtain significant information embedded in medical images, many deep learning-based algorithms have been proposed recently and achieved good fusion results. However, most of them can hardly capture the independent and underlying features, which leads to unsatisfactory fusion results. To address these issues, a multibranch residual attention reconstruction network (MBRARN) is proposed for the medical image fusion task. The proposed network mainly consists of three parts: feature extraction, feature fusion, and feature reconstruction. Firstly, the input medical images are converted into three scales by image pyramid operation and then are input into three branches of the proposed network respectively. The purpose of this procedure is to capture the local detailed information and the global structural information. Then, convolutions with residual attention modules are designed, which can not only enhance the captured outstanding features, but also make the network converge fast and stably. Finally, feature fusion is performed with the designed fusion strategy. In this step, a new more effective fusion strategy is correspondently designed for MRI-SPECT based on the Euclidean norm, called feature distance ratio (FDR). The experimental results conducted on Harvard whole brain atlas dataset demonstrate that the proposed network can achieve better results in terms of both subjective and objective evaluation, compared with some state-of-the-art medical image fusion algorithms. [ABSTRACT FROM AUTHOR]

In view of the matching problem between objects and mirrors in images, the RNFA(Relative Number of False Alarms)edge chain detection method is introduced to obtain richer line segments. An improved LBD(Line Band Descriptor) algorithm is proposed for constructing local invariant feature descriptors, and initial matching pairs are obtained by comparing local invariant feature descriptors. The screening of the global projection angle is adopted and the false matches in the initial matching pairs are eliminated fitting the projection center line. After the selection of the global projection angle and the fitting of the projection median are completed, the screening of the local invariant feature descriptor threshold is relaxed to obtain more matching pairs and improve the recall rate. The experimental results of image set simulation show that the designed algorithm can better identify line segments in the weaker texture regions and obtain more matching pairs on the basis of the guaranteed performance of the original algorithm, which can improve the correct matching rate by about 5% and achieve a recall rate of over 90%. [ABSTRACT FROM AUTHOR]

Copyright of Journal of South China University of Technology (Natural Science Edition) is the property of South China University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The accurate identification of highly similar sheet metal parts remains a challenging issue in sheet metal production. To solve this problem, this paper proposes an effective mean square differences (EMSD) algorithm that can effectively distinguish highly similar parts with high accuracy. First, multi-level downsampling and rotation searching are adopted to construct an image pyramid. Then, non-maximum suppression is utilised to determine the optimal rotation for each layer. In the matching, by re-evaluating the contribution of the difference between the corresponding pixels, the matching weight is determined according to the correlation between the grey value information of the matching pixels, and then the effective matching coefficient is determined. Finally, the proposed effective matching coefficient is adopted to obtain the final matching result. The results illustrate that this algorithm exhibits a strong discriminative ability for highly similar parts, with an accuracy of 97.1%, which is 11.5% higher than that of the traditional methods. It has excellent potential for application and can significantly improve sheet metal production efficiency. [ABSTRACT FROM AUTHOR]

We present a simple but effective technique to smooth out textures while preserving the prominent structures. Our method is built upon a key observation---the coarsest level in a Gaussian pyramid often naturally eliminates textures and summarizes the main image structures. This inspires our central idea for texture filtering, which is to progressively upsample the very low-resolution coarsest Gaussian pyramid level to a full-resolution texture smoothing result with well-preserved structures, under the guidance of each fine-scale Gaussian pyramid level and its associated Laplacian pyramid level. We show that our approach is effective to separate structure from texture of different scales, local contrasts, and forms, without degrading structures or introducing visual artifacts. We also demonstrate the applicability of our method on various applications including detail enhancement, image abstraction, HDR tone mapping, inverse halftoning, and LDR image enhancement. Code is available at https://rewindl.github.io/pyramid_texture_filtering/. [ABSTRACT FROM AUTHOR]

Deep learning models with convolutional operators have received widespread attention for their good image denoising performance. However, since the convolutional operation prefers to extract local features, the extracted features may lose some global information, such as texture, structure, and color characteristics, when the object in the image is large. To address this issue, this paper proposes an N-shaped convolutional neural network with the ability to extract multi-scale features to capture more useful information and alleviate the problem of global information loss. The proposed network has two main parts: a multi-scale input layer and a multi-scale feature extraction layer. The former uses a two-dimensional Haar wavelet to create an image pyramid, which contains the corrupted image's high- and low-frequency components at different scales. The latter uses a U-shaped convolutional network to extract features at different scales from this image pyramid. The method sets the mean-squared error as the loss function and uses the residual learning strategy to learn the image noise directly. Compared with some existing image denoising methods, the proposed method shows good performance in gray and color image denoising, especially in textures and contours. [ABSTRACT FROM AUTHOR]

The advent of deep learning in the past decade has significantly helped advance image inpainting. Although achieving promising performance, deep learning-based inpainting algorithms still struggle from the distortion caused by the fusion of structural and contextual features, which are commonly obtained from, respectively, deep and shallow layers of a convolutional encoder. Motivated by this observation, we propose a novel progressive inpainting network that maintains the structural and contextual integrity of a processed image. More specifically, inspired by the Gaussian and Laplacian pyramids, the core of the proposed network is a feature extraction module named GLE. Stacking GLE modules enables the network to extract image features from different image frequency components. This ability is important to maintain structural and contextual integrity, for high frequency components correspond to structural information while low frequency components correspond to contextual information. The proposed network utilizes the GLE features to progressively fill in missing regions in a corrupted image in an iterative manner. Our benchmarking experiments demonstrate that the proposed method achieves clear improvement in performance over many state-of-the-art inpainting algorithms.

This paper proposes an improved zero-mean normalization sum of squared differences (ZNSSD) algorithm to solve the problem of the inability of traditional structural measurement to extract high-frequency vibration signals. In the proposed technique, the high-speed image sequence of target vibration is captured by a high-speed camera. Then, the ZNSSD template matching algorithm with subpixel accuracy is introduced to process the captured images in the computer. Additionally, a modified search algorithm, the ZNSSD template matching algorithm based on image pyramid (ZNSSD-P), is proposed to significantly reduce the computation time and increase efficiency. Then, a jumping ZNSSD template matching algorithm based on image pyramid (J-ZNSSD-P) is proposed to further improve the efficiency of the ZNSSD-P algorithm. Vibration signals were extracted with Grating Ruler Motion Platform and sound barriers. Results show that the vibration signal extraction method has high precision and efficiency.

We have been participating in the Tsukuba Challenge, an open experiment involving autonomous robots, since 2014. The technology of our robot has stabilized, and our robot has continued to win the Tsukuba Mayor Prize from 2018 to 2021 without changing the basic configuration of the body and navigation software. Here, we report the robot's structure as the project's current completed form. Our robot is designed with the policy of selecting the most rational technology (appropriate technology) to achieve the purpose, even if it is not the latest. For example, we used image-like two-dimensional data instead of a three-dimensional point cloud in map matching for robot positioning. For pedestrian signal recognition, which was required to perform an optional task, we did not use deep learning but rather conventional color image processing. These techniques are advantageous for balancing the execution time and accuracy required in the challenge. [ABSTRACT FROM AUTHOR]

Citrus is an important commercial crop in many areas. The management and planning of citrus growing can be supported by timely and efficient monitoring of citrus-growing regions. Their complex planting structure and the weather are likely to cause problems for extracting citrus-growing regions from remote sensing images. To accurately extract citrus-growing regions, deep learning is employed, because it has a strong feature representation ability and can obtain rich semantic information. A novel model for extracting citrus-growing regions by UNet that incorporates an image pyramid structure is proposed on the basis of the Sentinel-2 satellite imagery. A pyramid-structured encoder, a decoder, and multiscale skip connections are the three main components of the model. Additionally, atrous spatial pyramid pooling is used to prevent information loss and improve the ability to learn spatial features. The experimental results show that the proposed model has the best performance, with the precision, the intersection over union, the recall, and the F1-score reaching 88.96%, 73.22%, 80.55%, and 84.54%, respectively. The extracted citrus-growing regions have regular boundaries and complete parcels. Furthermore, the proposed model has greater overall accuracy, kappa, producer accuracy, and user accuracy than the object-oriented random forest algorithm that is widely applied in various fields. Overall, the proposed method shows a better generalization ability, higher robustness, greater accuracy, and less fragmented extraction results. This research can support the rapid and accurate mapping of large-scale citrus-growing regions.

The aim of infrared (IR) and visible image fusion is to generate a more informative image for human observation or some other computer vision tasks. The activity-level measurement and weight assignment are two key parts in image fusion. In this paper, we propose a novel IR and visible fusion method based on the principal component analysis network (PCANet) and an image pyramid. Firstly, we use the lightweight deep learning network, a PCANet, to obtain the activity-level measurement and weight assignment of IR and visible images. The activity-level measurement obtained by the PCANet has a stronger representation ability for focusing on IR target perception and visible detail description. Secondly, the weights and the source images are decomposed into multiple scales by the image pyramid, and the weighted-average fusion rule is applied at each scale. Finally, the fused image is obtained by reconstruction. The effectiveness of the proposed algorithm was verified by two datasets with more than eighty pairs of test images in total. Compared with nineteen representative methods, the experimental results demonstrate that the proposed method can achieve the state-of-the-art results in both visual quality and objective evaluation metrics. [ABSTRACT FROM AUTHOR]

Information from an image occurs over multiple and distinct spatial scales. Image pyramid multiresolution representations are a useful data structure for image analysis and manipulation over a spectrum of spatial scales. This paper employs the Gaussian–Laplacian pyramid to separately treat different spatial frequency bands of a texture. First, we generate three images corresponding to three levels of the Gaussian–Laplacian pyramid for an input image to capture intrinsic details. Then, we aggregate features extracted from gray and color texture images using bioinspired texture descriptors, information-theoretic measures, gray-level co-occurrence matrix feature descriptors, and Haralick statistical feature descriptors into a single feature vector. Such an aggregation aims at producing features that characterize textures to their maximum extent, unlike employing each descriptor separately, which may lose some relevant textural information and reduce the classification performance. The experimental results on texture and histopathologic image datasets have shown the advantages of the proposed method compared to state-of-the-art approaches. Such findings emphasize the importance of multiscale image analysis and corroborate that the descriptors mentioned above are complementary. [ABSTRACT FROM AUTHOR]

Plant diseases have received common attention, and deep learning has also been applied to plant diseases. Deep neural networks (DNNs) have achieved outstanding results in plant diseases. Furthermore, DNNs are very fragile, and adversarial attacks in image classification deserve much attention. It is important to detect the robustness of DNNs through adversarial attacks. The paper firstly improves the EfficientNet by adding the SimAM attention module. The SimAM-EfficientNet is proposed in this paper. The experimental results show that the accuracy of the improved model on PlantVillage reaches 99.31%. The accuracy of ResNet50 is 98.33%. The accuracy of ResNet18 is 98.31%. The accuracy of DenseNet is 98.90%. In addition, the GP-MI-FGSM adversarial attack algorithm improved by gamma correction and image pyramid in this paper can increase the success rate of attack. The model proposed in this paper has an error rate of 87.6% whenattacked by the GP-MI-FGSM adversarial attack algorithm. The success rate of GP-MI-FGSM proposed in this paper is higher than other adversarial attack algorithms, including FGSM, I-FGSM, and MI-FGSM. [ABSTRACT FROM AUTHOR]

This paper proposes an improved zero-mean normalization sum of squared differences (ZNSSD) algorithm to solve the problem of the inability of traditional structural measurement to extract high-frequency vibration signals. In the proposed technique, the high-speed image sequence of target vibration is captured by a high-speed camera. Then, the ZNSSD template matching algorithm with subpixel accuracy is introduced to process the captured images in the computer. Additionally, a modified search algorithm, the ZNSSD template matching algorithm based on image pyramid (ZNSSD-P), is proposed to significantly reduce the computation time and increase efficiency. Then, a jumping ZNSSD template matching algorithm based on image pyramid (J-ZNSSD-P) is proposed to further improve the efficiency of the ZNSSD-P algorithm. Vibration signals were extracted with Grating Ruler Motion Platform and sound barriers. Results show that the vibration signal extraction method has high precision and efficiency. [ABSTRACT FROM AUTHOR]

Small object detection has always been a difficult direction in the field of object detection, especially the detection of small objects in UAV aerial images. The images captured by UAVs have the characteristics of small objects and dense objects. In order to solve these two problems, this paper improves the performance of object detection from the aspects of data and network structure. In terms of data, the data augmentation strategy and image pyramid mechanism are mainly used. The data augmentation strategy adopts the method of image division, which can greatly increase the number of small objects, making it easier for the algorithm to be fully trained during the training process. Since the object is denser, the image pyramid mechanism is used. During the training process, the divided images are up-sampled into three different sizes, and then sent to three different detectors respectively. Finally, the detection results of the three detectors are fused to obtain the final detection results. The small object itself has few pixels and few features. In order to improve the detection performance, it is necessary to use context. This paper adds attention mechanism to the yolov5 network structure, while adding a detection head to the underlying feature map to make the network structure pay more attention to small objects. By using data augmentation and improved network structure, the detection performance of small objects can be significantly improved. The experiment in this paper is carried out on the Visdrone2019 dataset and DOTA dataset. Through experimental verification, our proposed method can significantly improve the performance of small object detection. [ABSTRACT FROM AUTHOR]

The accurate identification of highly similar sheet metal parts remains a challenging issue in sheet metal production. To solve this problem, this paper proposes an effective mean square differences (EMSD) algorithm that can effectively distinguish highly similar parts with high accuracy. First, multi-level downsampling and rotation searching are adopted to construct an image pyramid. Then, non-maximum suppression is utilised to determine the optimal rotation for each layer. In the matching, by re-evaluating the contribution of the difference between the corresponding pixels, the matching weight is determined according to the correlation between the grey value information of the matching pixels, and then the effective matching coefficient is determined. Finally, the proposed effective matching coefficient is adopted to obtain the final matching result. The results illustrate that this algorithm exhibits a strong discriminative ability for highly similar parts, with an accuracy of 97.1%, which is 11.5% higher than that of the traditional methods. It has excellent potential for application and can significantly improve sheet metal production efficiency.

Deep learning models with convolutional operators have received widespread attention for their good image denoising performance. However, since the convolutional operation prefers to extract local features, the extracted features may lose some global information, such as texture, structure, and color characteristics, when the object in the image is large. To address this issue, this paper proposes an N-shaped convolutional neural network with the ability to extract multi-scale features to capture more useful information and alleviate the problem of global information loss. The proposed network has two main parts: a multi-scale input layer and a multi-scale feature extraction layer. The former uses a two-dimensional Haar wavelet to create an image pyramid, which contains the corrupted image’s high- and low-frequency components at different scales. The latter uses a U-shaped convolutional network to extract features at different scales from this image pyramid. The method sets the mean-squared error as the loss function and uses the residual learning strategy to learn the image noise directly. Compared with some existing image denoising methods, the proposed method shows good performance in gray and color image denoising, especially in textures and contours.

Pedestrian detection (PD) systems capable of locating pedestrians over large distances and locating them faster are needed in Pedestrian Collision Prediction (PCP) systems to increase the decision-making distance. This paper proposes a performance-optimized FPGA implementation of a HOG-SVM-based PD system with support for image pyramids and detection windows of different sizes to locate near and far pedestrians. This work proposes a hardware architecture that can process one pixel per clock cycle by exploring data and temporal parallelism using techniques such as pipeline and spatial division of data between parallel processing units. The proposed architecture for the PD module was validated in FPGA and integrated with the stereo semi-global matching (SGM) module, also prototyped in FPGA. Processing two windows of different dimensions permitted a reduction in miss rate of at least 6% compared to a uniquely sized window detector. The performances achieved by the PD system and the PCP system in HD resolution were 100 and 66.2 frames per second (FPS), respectively. The performance improvement achieved by the PCP system with the addition of our PD module permitted an increase in decision-making distance of 3.3 m compared to a PCP system that processes at 30 FPS. [ABSTRACT FROM AUTHOR]

In this paper, a simple feature fusion scheme is proposed for content-based image retrieval (CBIR) using color, texture and shape feature moments. The low dimensional feature descriptor is constructed by fusing color, texture and shape moments of an image effectively. The color moments are extracted from the image using probability histogram model while the Gray Level Co-occurrence Matrix (GLCM) based texture moments are computed in a very new fashion by selecting salient components in Discrete Cosine Transform (DCT) domain after determining the inter-relationship between the DCT blocks. Alone, color or texture information is not adequate to produce the desire results in image retrieval system. So, suggested work has also considered the multi-resolution based shape information, since the single resolution level of image does not provide an adequate image information and it may be possible that the fine details may be visualized on other resolution levels of image. Therefore, shape feature descriptor is determined by calculating the invariant moments of multi-resolution based sub-images at the various levels. Finally, fused single feature descriptor is computed by proficient fusion of color, texture and shape feature moments. The modified distance is also suggested for image retrieval task. The proposed feature fusion scheme is implemented on Corel-1K, OT-8 and GHIM-10K image databases to evaluate the retrieval performance and experimental results show the effectiveness of our scheme over the other existing CBIR schemes. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Harbin University of Science & Technology is the property of Journal of Harbin University of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this paper, a distributed, autonomous, cooperative mission-planning (DACMP) approach was proposed to focus on the problem of the real-time cooperative searching and surveillance of multiple unmanned aerial vehicles (multi-UAVs) with threats in uncertain and highly dynamic environments. To deal with this problem, a time-varying probabilistic grid graph was designed to represent the perception of a target based on its a priori dynamics. A heuristic search strategy based on pyramidal maps was also proposed. Using map information at different scales makes it easier to integrate local and global information, thereby improving the search capability of UAVs, which has not been previously considered. Moreover, we proposed an adaptive distributed task assignment method for cooperative search and surveillance tasks by considering the UAV motion environment as a potential field and modeling the effects of uncertain maps and targets on candidate solutions through potential field values. The results highlight the advantages of search task execution efficiency. In addition, simulations of different scenarios show that the proposed approach can provide a feasible solution for multiple UAVs in different situations and is flexible and stable in time-sensitive environments.

Deep learning (DL)-based methods have recently been extensively used for satellite image analysis due to their ability to automatically extract spatial-spectral features from images. Recent advancement in DL-based methods has also allowed the remote sensing community to utilize these methods in fusing the satellite images for enhanced land use/land cover (LULC) classification. Keeping this in view, the present study aims to evaluate the potential of SAR (Sentinel 1) and Optical (Sentinel 2) image fusion using pyramid-based DL methods over an agricultural area in India. In this study, three image fusion methods, i.e., pyramid-based fusion methods, pyramid-based fusion methods coupled with convolutional neural network (CNN), and a combination of two different pyramid decomposition methods concurrently with CNN were used. The performance of the fused images was evaluated in terms of fusion metrics, image quality, and overall classification accuracy by an optimized 2D-CNN-based DL classifier. Results from pyramid-based fusion methods with CNN and a combination of two different pyramid decomposition methods with CNN suggest that these methods were able to retain visual quality as well as the detailed structural information of input images in comparison to the pyramid-based fusion methods. Bayesian optimization method was used to optimize various hyper-parameters of the 2D-CNN-based DL classifier used in this study. Results with fused images obtained using pyramid-based methods coupled with CNN suggest an improved performance by VV (Vertical–Vertical) polarized images in terms of overall classification accuracy (99.23% and 99.33%). [ABSTRACT FROM AUTHOR]

To extract craters with a radius greater than 10 km more effectively from lunar digital elevation maps, pyramidal image segmentation based on the U-Net model is proposed, and the conversion relationship between the multilayer image pyramid and the geographic coordinates of the crater is established. The crater image pyramid method ensures the full coverage of the study area with a small number of images and that each crater exists in several images with different resolutions. The proposed method can effectively improve the detection performance of large-scale craters and solve the migration problem when stitching together craters from large-scale images. This method recovered 85.48% of the craters with a radius greater than 10 km in an artificially annotated dataset, found 1044 new craters, and extended the maximum radius of detected craters from 72 km in randomly cropped image segmentation to 200 km. It was estimated by visual interpretation that approximately 82.09% of these new craters are real. Also, the recall reaches 90.17% when the new real craters are added to the true craters. [ABSTRACT FROM AUTHOR]

The current color difference online detection algorithm has poor real-time performance. The noise affects color difference monitoring and sampling detection results are different from online detection. Therefore, a color difference online detection algorithm is studied. Firstly, the Flat Field and Perfect Reflector Method (PRM) are used to correct the color illumination uniformity of warp knitted fabric images. The image pyramid principle is used to down-sample the images to improve the online detection speed. Secondly, the principle of the histogram intersection method is used to pre-judge the color difference of warp knitted fabric image, and judge whether there is a color difference. The improved K-means algorithm is used to analyze the color types of the fabric, and the CMC (l : c) calculation formula is used to quantitatively calculate the color difference value. Finally, the algorithm has been verified in the factory. The speed of fabric inspection using the algorithm can reach 0.75–0.8m/s, and the color difference detection algorithm is compared with manual detection. The results show that the method can meet the real-time requirements of the system and it is consistent with the sampling test, and the detection speed increased from 0.2 m / s to 0.8 m / s. [ABSTRACT FROM AUTHOR]

Image block matching is one of active research fields in image processing, which has been widely used in security monitoring and motion estimation. Due to great disparities in images of the same scene under various illumination, block matching has been a challenging task. To this end, we propose a hierarchical block matching method which is adaptive to the computational complexity and suited to complicated illumination environments. The approach is divided into two parts. First, in order to reduce searching time, the whole algorithm is established in the framework of pyramid structures. Second, the correlation coefficient and structural functions are adopted to evaluate the similarity between two images so as to get better matching results. Simulation results show that, compared with the classical three-step, four-step searching algorithms and OHBM algorithm, this algorithm can reduce the time complexity efficiently. Moreover, since the algorithm makes use of the structure details of images, it is robust to complicated scenes. [ABSTRACT FROM AUTHOR]

Additive manufacturing (AM) is a type of advanced manufacturing process that enables fast prototyping to realize personalized products in complex shapes. However, quality defects existed in AM products can directly lead to significant failures (e.g., cracking caused by voids) in practice. Thus, various inspection techniques have been investigated to evaluate the quality of AM products, where X-ray computed tomography (CT) serves as one of the most accurate techniques to detect geometric defects (e.g., voids inside an AM product). Taking a selective laser melting (SLM) process as an example, voids can be detected by investigating CT images after the fabrication of products with limited disturbance from noises. However, limited by the sensor size and scanning speed issue, CT is difficult to be used for online (i.e., layer-wise) voids detection, monitoring, and process control to mitigate the defects. As an alternative, optical cameras can provide layer-wise images to support online voids detection. The intricate texture of the layer-wise image restricts the accuracy of void detection in AM products. Therefore, we propose a new method called pyramid ensemble convolutional neural network (PECNN) to efficiently detect voids and predict the texture of CT images using layer-wise optical images. The proposed PECNN can efficiently extract informative features based on the ensemble of the multiscale feature-maps (i.e., image pyramid) from optical images. Unlike deterministic ensemble strategies, this ensemble strategy is optimized by training a neural network in a data-driven manner to learn the fine-grained information from the extracted feature-maps. The merits of the proposed method are illustrated by both simulations and a real case study in a SLM process. [ABSTRACT FROM AUTHOR]

The Friction stir welding process is a new entrant in welding technology. The FSW joints have high strength and helps in weight saving considerably than the other joining process as no filler material is added during welding. The weld quality is affected because of various kinds of defects occurring during the FSW process. Defects like cavity, surface grooves and flash could occur due to inappropriate set of process parameters which results in excessive or insufficient heat input. Defects analysis can be done by several non-destructive methods like immersion ultrasonic techniques, X-ray radiography, thermography, eddy current testing, synchrotron technique etc. In the present work the image processing techniques are applied over the test samples to detect the surface defects like pin holes, surface grooves etc.

Multi-scale exposure fusion is a powerful approach to fuse variously low dynamic range images into a high quality image. Fine and attractive information of source images is added in the resultant image. It may yield better fusion result than the images fused by single scale exposure fusion. In multi-scale exposure fusion possibility of halo artifact can be produced and the details in darkest and brightest areas are normally not retained in the final fused image. Based on image segmentation and edge preserving filtration technique a novel algorithm is proposed in this paper for multi-scale exposure images. Taking advantage of super pixel (image segmentation) along with edge preserving technique, details in the darkest and brightest regions are well protected which removes the halo artifacts produced in the fused image. The outcome of experiments proved that proposed algorithm provides better results than other state of the art algorithms used for image fusion.

Traditional face detection algorithms often cannot extract useful detection features from the original image, and convolutional neural networks can easily extract high-dimensional feature information, which is widely used in image processing. In view of the above shortcomings, a simple and efficient deep learning Caffe framework is adopted and trained by AlexNet network. The data set is LFW face dataset, and a model classifier is obtained. Image pyramid transformation is performed on the original image data, and feature graph is obtained by forward propagation. The inverse transformation yields the face coordinates, uses the non-maximum suppression algorithm to obtain the optimal position, and finally reaches a two-class face detection result. The method can realize face detection with different scales and has high precision, and can be used to construct a face detection system.

Information from an image occurs over multiple and distinct spatial scales. Image pyramid multiresolution representations are a useful data structure for image analysis and manipulation over a spectrum of spatial scales. This paper employs the Gaussian–Laplacian pyramid to separately treat different spatial frequency bands of a texture. First, we generate three images corresponding to three levels of the Gaussian–Laplacian pyramid for an input image to capture intrinsic details. Then, we aggregate features extracted from gray and color texture images using bioinspired texture descriptors, information-theoretic measures, gray-level co-occurrence matrix feature descriptors, and Haralick statistical feature descriptors into a single feature vector. Such an aggregation aims at producing features that characterize textures to their maximum extent, unlike employing each descriptor separately, which may lose some relevant textural information and reduce the classification performance. The experimental results on texture and histopathologic image datasets have shown the advantages of the proposed method compared to state-of-the-art approaches. Such findings emphasize the importance of multiscale image analysis and corroborate that the descriptors mentioned above are complementary.

The aim of infrared (IR) and visible image fusion is to generate a more informative image for human observation or some other computer vision tasks. The activity-level measurement and weight assignment are two key parts in image fusion. In this paper, we propose a novel IR and visible fusion method based on the principal component analysis network (PCANet) and an image pyramid. Firstly, we use the lightweight deep learning network, a PCANet, to obtain the activity-level measurement and weight assignment of IR and visible images. The activity-level measurement obtained by the PCANet has a stronger representation ability for focusing on IR target perception and visible detail description. Secondly, the weights and the source images are decomposed into multiple scales by the image pyramid, and the weighted-average fusion rule is applied at each scale. Finally, the fused image is obtained by reconstruction. The effectiveness of the proposed algorithm was verified by two datasets with more than eighty pairs of test images in total. Compared with nineteen representative methods, the experimental results demonstrate that the proposed method can achieve the state-of-the-art results in both visual quality and objective evaluation metrics.

Infrared and visible image fusion is an effective image processing technique to obtain more comprehensive information, which can help people better understand various scenarios. In this paper, a novel infrared and visible image fusion method is proposed which fully considers the attributes of objects in source images. Benefiting from the attribute and the edge-preserving filters, the prominent objects in the infrared source image are effectively extracted. Then, the weight-based Laplacian pyramid fusion strategy is adopted to get more natural fusion results. The experimental results on the public image fusion datasets and a new infrared–visible video fusion dataset show that the proposed method achieves state-of-the-art fusion performance in terms of both visual and objective evaluations. The proposed algorithm is also implemented in an infrared–visible dual sensor system, which demonstrated the practicability of our fusion method. • A novel attribute guided image fusion method is proposed. • A dual infrared and visible imaging system is constructed. • A high-quality infrared–color visible image and video dataset is established. [ABSTRACT FROM AUTHOR]

The fast and precise positioning of lithium battery is crucial for effective manufacturing of mass production. In order to acquire position information of lithium batteries rapidly and accurately, a novel dual-template matching algorithm is proposed to properly locate and segment each battery for fast and precise mass production. Initially, an image down-sampling method is applied to build up a multi-layer image pyramid for speeding up target search, and a novel mixed matching template is designed to increase the matching precision. A row of lithium batteries is likely tilt during rolling, and the images of batteries captured by the CCD camera are distorted, which may generate a negative effect on next procedure. Hence, a two-level correction algorithm for battery angle and location is applied to obtain rough areas of the batteries and improve the accuracy of template matching. Lastly, the comparison with other state-of-the-art algorithms is done to locate each battery in a row with high speed and precision. The precision rates of the proposed algorithm, improved SAD algorithm, and YOLOv3 algorithm are 99.44%, 95.98%, and 93.64 for normal battery images and 97.86%, 89.19%, and 85.10 for tilted battery images, respectively. Compared with improved SAD matching algorithm and YOLOv3 algorithm, the positioning accuracy of the proposed method is significantly increased, and the matching robustness is improved in spite of large battery inclination angle. [ABSTRACT FROM AUTHOR]

Small object detection has always been a difficult direction in the field of object detection, especially the detection of small objects in UAV aerial images. The images captured by UAVs have the characteristics of small objects and dense objects. In order to solve these two problems, this paper improves the performance of object detection from the aspects of data and network structure. In terms of data, the data augmentation strategy and image pyramid mechanism are mainly used. The data augmentation strategy adopts the method of image division, which can greatly increase the number of small objects, making it easier for the algorithm to be fully trained during the training process. Since the object is denser, the image pyramid mechanism is used. During the training process, the divided images are up-sampled into three different sizes, and then sent to three different detectors respectively. Finally, the detection results of the three detectors are fused to obtain the final detection results. The small object itself has few pixels and few features. In order to improve the detection performance, it is necessary to use context. This paper adds attention mechanism to the yolov5 network structure, while adding a detection head to the underlying feature map to make the network structure pay more attention to small objects. By using data augmentation and improved network structure, the detection performance of small objects can be significantly improved. The experiment in this paper is carried out on the Visdrone2019 dataset and DOTA dataset. Through experimental verification, our proposed method can significantly improve the performance of small object detection.