Heterogeneous synthetic aperture radar (SAR) data contain more information, so the use of heterogeneous SAR images can potentially improve the performance of remote sensing applications. Feature tracking is crucial for using heterogeneous SAR data. However, feature tracking is a challenge using heterogeneous SAR images to harmonize high-resolution (HR) data with coarser data. In this paper, we propose a smooth edge-guide super-resolution recurrent residual learning network to uniform resolution of heterogeneous SAR image such that their features have more consistent representation. Our proposed framework contains a super-resolution network that aims to translate the low-resolution (LR) images into the HR ones to reduce their feature differences. The generated HR final image and the HR raw image can be considered homogeneous for sea ice drift tracking. Through several examples, we demonstrate the effectiveness of the method in the feature matching of images with a large-resolution difference images from different SAR sensors.

Abstract In medical image segmentation, the neural network structure of the U‐Net family has demonstrated sufficient advantages. However, MRI images have different scan parameters and different scan times, resulting in different feature representation of the images. Furthermore, there is a great class imbalance between bone and cartilage tissues in MRI knee images. To address these issues, a region‐based two‐stage MRI knee bone tissue segmentation network is proposed in this paper. The segmentation network makes full use of the location characteristics of the three types of bone tissue in the knee joint and uses a two‐stage network architecture with a modified U 2 ‐Net backbone network to segment MRI knee bone tissue. The neural network structure is divided into two phases, the first phase with a simple coded decoding structure for saliency detection to obtain the positional regional relationships of different bone tissues, and the second phase with a segmentation network consisting of 2 modified U 2 ‐Net, one for segmenting the patella and associated cartilage and the other for segmenting the femur, tibia and associated cartilage. The algorithm was tested with a variety of MRI knee data to verify the effectiveness of the algorithm.

GF-1 multispectral wide field of view (WFV) images, with a spatial resolution of 16 m, have been widely used in earth monitoring. However, the spatial details provided by WFV images are not sufficient for many applications. Thus, this letter proposes a novel WFV image super-resolution (SR) algorithm called GFRCAN based on a very deep residual coordinate attention network. To form a very deep network, the residual-in-residual (RIR) structure consisting of several residual groups (RG) with long skip connections is used. Meanwhile, the residual coordinate attention block (RCOAB) and adaptive multi-scale spatial attention module (AMSA) are incorporated to focus on the high-frequency information and multi-scale features adaptive weighted fusion. Besides, the spectral and spatial details of SR images are improved by incorporating peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) into the loss function. Both subjective and objective evaluation results show that the proposed model outperforms the state-of-the-art methods.

Synthetic aperture radar automatic target recognition (SAR ATR) is a key technique of remote-sensing image recognition, which has many potential applications in the fields of military surveillance, national defense, civil application, and so on. With the development of science and technology, deep convolutional neural network (DCNN) has been widely applied for SAR ATR. However, it is difficult to use deep learning to train models with limited ray SAR images. To resolve this problem, we proposed an effectively lightweight attention mechanism CNN (AM-CNN) model for SAR ATR. Extensive experimental results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set illustrate that the AM-CNN model can achieve a superior recognition performance, and the average recognition accuracy can reach 99.35% on the classification of 10 class targets. Compared with the traditional CNN and the state-of-the-art method, our model is significantly superior to improve performance and efficiency.

In the above article[1], there was a typographical error in Fig. 2. The correct figure appears below:

Accurate point matching is widely used, and it is a critical and challenging process in feature-based image registration. To improve feature matching accuracy on putative matches with heavy outliers and similar local structures, an accurate and robust feature point matching algorithm based on minimum relative motion entropy (MRME) is proposed, in which the relative motion between the putative matches and their K-nearest neighbors is formulated. Based on the relative motion clustering result, the relative motion entropy is defined to find the coincident relative motions. According to relative motions with MRME, the outliers are removed in a two-stage feature match strategy. With quasi-linear time complexity, outliers with random or irregular relative motion are removed efficiently and accurately, while inliers with coincident relative motion are retained. Three data sets with repetitive patterns, viewpoint changes, low overlapping areas, and local deformations are used to demonstrate the performance of the proposed algorithm. MRME is shown to be more robust and accurate than ten state-of-the-art feature matching algorithms.

Existing methods used in detecting vehicles with traffic violations are mostly based on single-entity frameworks, and those involving multientities remain to be limited. In this paper, we propose a traffic violation detection model, an intelligent vehicle violation recognition method based on multiple entities. It aims to identify vehicles violating the rule of yielding to pedestrians at nonsignalized crosswalks. First, we define the concepts of vehicles and pedestrians and then apply the regression background subtraction method and particle filter algorithm to automatically identify and track moving objects. Second, the pedestrian validity feature template is specified to detect temporal trajectory features from videos with labels and to train classification networks aiming to identify unreasonable behavior patterns, such as remaining on the sidewalk, entering the roadway, bicycle riding, and others. Finally, we develop a novel traffic violation recognition method based on multientity interaction analysis. The cases of failing to yield to pedestrians are recognized based on the multientity feature template built using the proposed method. We verified the effectiveness of the proposed method on a real traffic data set obtained from surveillance cameras. The obtained results are significantly better compared with the baseline method. The area under curve value of the proposed traffic violation recognition method with multientity interaction is 14.5%, 11.1%, and 6.6% higher compared with the three baseline methods based on single-entity frameworks.

Due to the existence of repetitive patterns, ambiguous features, and similar local structures in remote sensing images, it is inevitable that the outliers with local pseudoisomorphic structures are preserved as inliers, which makes point matching still a challenging problem. To improve the accuracy of feature matching, a discriminative point matching algorithm named local structure consensus constraint is proposed to remove the outliers from putative correspondences and find two local structure consensus graphs composed of inliers. First, a local structure descriptor is proposed to evaluate the corresponding structure similarity of the K-nearest neighbors. Then, a cost function is defined to evaluate the local structure consistency. With a two-stage outlier removing strategy, the feature points with different local structures are eliminated, and two local structure consensus graphs are obtained. To evaluate the performance of the proposed algorithm, 45 aerial image pairs taken around the Shandong Peninsula with repetitive local patterns and ambiguous features are used. Compared with five state-of-the-art point matching methods, the proposed algorithm is proven to be more accurate and efficient.

The phenomenon of spectral aliasing exists for coastal wetland object types, which leads to class mixing. This letter proposes a multiobject convolutional neural network (CNN) decision fusion classification method for hyperspectral images of coastal wetlands. This method adopts decision fusion based on fuzzy membership rules applied to single-object CNN classification to obtain higher classification accuracy. Experimental results demonstrate the effectiveness of the proposed method for the six object types, including water, tidal flat, reed, and other vegetation types. The overall accuracy of the decision fusion classification method based on fuzzy membership is 82.11%, which is 3.33% and 6.24% higher than those of single-object feature band CNN and support vector machine methods. The classification method based on multiobject CNN decision fusion inherits the characteristics of single-object feature bands of the CNN, making it a practical approach to image classification under the challenging conditions in which class mixing occurs.

Nowadays, the accidents of oil spill become more and more frequent, causing pollution to the natural resources, marine environment and lives in the sea. As a result, the detection of oil spill draws more and more attentions. One of the most popular region-based active contour models proposed by Chan and Vese, is widely used to image segmentation. But it can't segment hyperspectral oil spill image well, which has blurry boundaries, low distinction, and noise and so on. In order to segment oil spill region from the hyperspectral oil spill image accurately, we improved the region-based active contour model in this paper. For the energy functional, we firstly bring the thought of Fisher criterion into the fitting term to get a better classification result faster. Secondly, a new stop function based on gradient of spectral angle measurement is added into the length term, so as to take advantage of the edge information fully even it is blurry. At last, the model is extended to be able to segment desired material from the complex image with several classes in it. We take some experiments on synthetic and real hyperspectral images to verify the effectiveness of our model, and apply it to the airborne hyperspectral oil spill image. Results of the proposed model on synthetic and testing hyperspectral images show that it outperforms the CV model greatly, and does better than several other segmentation and classification algorithms. Results on hyperspectral oil spill images show that it improves the ability of distinguishing oil spills from sea water, even there are boats and flats in the image.

The detection of targets with complex backgrounds in aerial images is a challenging task. In this letter, we propose a robust insulator detection algorithm based on local features and spatial orders for aerial images. First, we detect local features and introduce a multiscale and multifeature descriptor to represent the local features. Then, we get several spatial orders features by training these local features, it improves the robustness of the algorithm. Finally, through a coarse-to-fine matching strategy, we eliminate background noise and determine the region of insulators. We test our method on a diverse aerial image set. The experimental results demonstrate the precision and robustness of our detection method, and indicate the possible use of our method in practical applications.

The objects in natural images are often texturally inhomogeneous and prone to be falsely segmented into different parts by conventional methods. To overcome the difficulties caused by texture inhomogeneity, a new active contour model is proposed to extract inhomogeneous insulators from aerial images. First, a semilocal operator is employed to extract the texture features of insulators under the Beltrami framework. The layer of semilocal texture feature is single, and thus, it can avoid the high dimensionality of feature space. Then, a new convex energy functional is defined by taking the Xie's nonconvex model into a global minimization active contour framework during the process of segmentation. The proposed energy functional consists of not only the semilocal texture features of insulators but also their spatial relationship, which improves its ability to deal with textural inhomogeneity. Moreover, it can also avoid the existence of local minima in the minimization of the Xie's nonconvex model, thereby being independent of initial contour. In the process of contour evolution and numerical minimization, a fast dual formulation is employed to overcome the drawbacks of the usual level set and gradient descent method and to make the evolution of the contour more efficient. The experimental results on aerial insulator images confirm the ability of the proposed algorithm to effectively segment inhomogeneous textures with an overall average rmse of 1.87 pixels, a precision of 85.59%, and a recall of 86.47%. In addition, the proposed algorithm is extended to animal images, and satisfactory segmentation results can be obtained as well.

Accurate point matching is a critical and challenging process in feature-based image registration. In this paper, a simple and robust feature point matching algorithm, called Restricted Spatial Order Constraints (RSOC), is proposed to remove outliers for registering aerial images with monotonous backgrounds, similar patterns, low overlapping areas, and large affine transformation. In RSOC, both local structure and global information are considered. Based on adjacent spatial order, an affine invariant descriptor is defined, and point matching is formulated as an optimization problem. A graph matching method is used to solve it and yields two matched graphs with a minimum global transformation error. In order to eliminate dubious matches, a filtering strategy is designed. The strategy integrates two-way spatial order constraints and two decision criteria restrictions, i.e., the stability and accuracy of transformation error. Twenty-nine pairs of optical and Synthetic Aperture Radar (SAR) aerial images are utilized to evaluate the performance. Compared with RANdom SAmple Consensus (RANSAC), Graph Transformation Matching (GTM), and Spatial Order Constraints (SOC), RSOC obtained the highest precision and stability.

Edge detection is a crucial approach for the location and acreage calculation of oil slick when oil spills on the sea. In this paper, in view of intensity inhomogeneity, high noise, and blurring of oil slick infrared (IR) aerial images, a novel algorithm is proposed to detect the edges of oil slick IR aerial images. In the proposed algorithm, we define an energy function model combining a region-scalable-fitting concept and a global minimization active contour (GMAC) model. The proposed novel algorithm avoids the existence of local minima and meanwhile deals with the intensity inhomogeneity, noise, and weak edge boundaries exiting in oil spill IR images. In the process of the active contour evolving toward object boundaries and numerical minimization, a dual formulation is used for overcoming drawbacks of the usual level set and gradient descent method so that the process of minimization can be much easier and our algorithm is independent of the initial position of the contour. Using the proposed algorithm, we can gain continuous and closed edges of oil slick IR aerial images. The experiment results have shown that the proposed algorithm outperforms conventional edge detection methods and other algorithms in terms of the efficiency and accuracy. In addition, the proposed algorithm is extended to synthetic-aperture-radar oil slick images, and satisfactory results of edge extraction can be obtained as well.