Feature Matching Results Research Articles

Reference-based super-resolution reconstruction of remote sensing images based on a coarse-to-fine feature matching transformer

Remote sensing image super-resolution reconstruction technology mines the deep details of remote sensing image, which has been widely used in the fields of intelligent and precise agriculture, intelligent transportation, earth surface object recognition, and so on. To obtain more detailed information,we design a reference-based super-resolution reconstruction network. Firstly, the coarse-to-fine feature matching strategy is adopted for the features of the input image. Global coarse matching is performed on the center patch of each block, and then pixel-level local fine matching is performed on the edge patches of the block. This both reduces the amount of computation and improves the matching accuracy. A threshold is set to determine whether the feature matching results meet the criteria for feature transfer. Finally, different scale features are fused through several convolutional layers and sampling operations, obtaining the reconstruction features after a fourfold increase in resolution. The ultimate super-resolution image is generated through a decoder. We have performed training and testing on remote sensing datasets. Compared to the current state-of-the-art methods, our proposed method is visually more details and outperforms other methods in terms of objective evaluation metrics.

Intelligent Vehicle Inspection Tool Design Based on Freeman Chain Code for Automatic Annotation of 3D Models

Autonomous vehicle are more and more widely used in daily life, and the requirements for their safety performance are higher and higher. As a tool for testing auto parts, intelligent inspection tools are crucial to the guarantee of automobile quality. However, traditional fixture design relies on manual drawing, which is inefficient and prone to errors. To solve this problem, this research uses Freeman chain code to determine the annotation object, uses case clustering method to annotate, and uses error back propagation algorithm to realize case knowledge classification learning, and designs intelligent vehicle inspection tool design technology based on Freeman chain code 3D automatic annotation method. The experimental results show that the geometric feature matching results are correct, and the difference in feature comparison results is significant, with a high accuracy rate. Meanwhile, the geometric similarity annotation method has a high accuracy rate, taking only 3 minutes to complete the annotation, which is 7 minutes longer than traditional manual annotation. The error backpropagation algorithm can accurately achieve feature classification, and the design time of size chain inspection tool deformation design is reduced by 214 minutes compared to manual reverse deformation design, significantly improving design efficiency. In summary, the proposed design method for automotive inspection tools can achieve automatic model annotation, improve design efficiency, and reduce design time.

Soft Tissue Feature Tracking Based on Deep Matching Network

Research in the field of medical image is an important part of the medical robot to operate human organs. A medical robot is the intersection of multi-disciplinary research fields, in which medical image is an important direction and has achieved fruitful results. In this paper, a method of soft tissue surface feature tracking based on a depth matching network is proposed. This method is described based on the triangular matching algorithm. First, we construct a self-made sample set for training the depth matching network from the first N frames of speckle matching data obtained by the triangle matching algorithm. The depth matching network is pre-trained on the ORL face data set and then trained on the self-made training set. After the training, the speckle matching is carried out in the subsequent frames to obtain the speckle matching matrix between the subsequent frames and the first frame. From this matrix, the inter-frame feature matching results can be obtained. In this way, the inter-frame speckle tracking is completed. On this basis, the results of this method are compared with the matching results based on the convolutional neural network. The experimental results show that the proposed method has higher matching accuracy. In particular, the accuracy of the MNIST handwritten data set has reached more than 90%.

Coarse TRVO: A Robust Visual Odometry with Detector-Free Local Feature

The visual SLAM system requires precise localization. To obtain consistent feature matching results, visual features acquired by neural networks are being increasingly used to replace traditional manual features in situations with weak texture, motion blur, or repeated patterns. However, to improve the level of accuracy, most deep learning enhanced SLAM systems, which have a decreased efficiency. In this paper, we propose Coarse TRVO, a visual odometry system that uses deep learning for feature matching. The deep learning network uses a CNN and transformer structures to provide dense high-quality end-to-end matches for a pair of images, even under indistinctive settings with low-texture regions or repeating patterns occupying the majority of the field of view. Meanwhile, we made the proposed model compatible with NVIDIA TensorRT runtime to boost the performance of the algorithm. After obtaining the matching point pairs, the camera pose is solved in an optimized way by minimizing the re-projection error of the feature points. Experiments based on multiple data sets and real environments show that Coarse TRVO achieves a higher robustness and relative positioning accuracy in comparison with the current mainstream visual SLAM system.

Feature matching for texture-less endoscopy images via superpixel vector field consistency.

Feature matching is an important technology to obtain the surface morphology of soft tissues in intraoperative endoscopy images. The extraction of features from clinical endoscopy images is a difficult problem, especially for texture-less images. The reduction of surface details makes the problem more challenging. We proposed an adaptive gradient-preserving method to improve the visual feature of texture-less images. For feature matching, we first constructed a spatial motion field by using the superpixel blocks and estimated its information entropy matching with the motion consistency algorithm to obtain the initial outlier feature screening. Second, we extended the superpixel spatial motion field to the vector field and constrained it with the vector feature to optimize the confidence of the initial matching set. Evaluations were implemented on public and undisclosed datasets. Our method increased by an order of magnitude in the three feature point extraction methods than the original image. In the public dataset, the accuracy and F1-score increased to 92.6% and 91.5%. The matching score was improved by 1.92%. In the undisclosed dataset, the reconstructed surface integrity of the proposed method was improved from 30% to 85%. Furthermore, we also presented the surface reconstruction result of differently sized images to validate the robustness of our method, which showed high-quality feature matching results. Overall, the experiment results proved the effectiveness of the proposed matching method. This demonstrates its capability to extract sufficient visual feature points and generate reliable feature matches for 3D reconstruction and meaningful applications in clinical.

A Unified Framework of Bundle Adjustment and Feature Matching for High-Resolution Satellite Images

Bundle adjustment (BA) is a technique for refining sensor orientations of satellite images, while adjustment accuracy is correlated with feature matching results. Feature matching often contains high uncertainties in weak/repeat textures, while BA results are helpful in reducing these uncertainties. To compute more accurate orientations, this article incorporates BA and feature matching in a unified framework and formulates the union as the optimization of a global energy function so that the solutions of the BA and feature matching are constrained with each other. To avoid a degeneracy in the optimization, we propose a comprised solution by breaking the optimization of the global energy function into two-step suboptimizations and compute the local minimums of each suboptimization in an incremental manner. Experiments on multi-view high-resolution satellite images show that our proposed method outperforms state-of-the-art orientation techniques with or without accurate least-squares matching.

Automatic assembly of micro-miniature parts based on coaxial alignment and ORB feature matching

Aiming at the problems of low manual assembly process efficiency and low yield of micro-miniature parts. A vision system based on ORB feature matching for fast coaxial alignment of micro-miniature parts automated assembly is proposed. The coaxial alignment module is the main hardware, the assembly part and the base part can be imaged in the industrial camera; the ORB features are extracted from the known part template image and the part image obtained by the vision imaging system, and the RANSAC algorithm is adopted to match the ORB features of the template image and the actual part image, the pose of parts is calculated by feature matching results. At the same time, the image coordinate system and the motion axis coordinate system are calibrated, and the motion control system is driven by the transformation matrix of the two coordinate systems to complete the assembly between the assembly part and the base part. The assembly experiment shows the system can complete the automated assembly of micro-miniature parts.

Underwater Image Dehazing via Unpaired Image-to-image Translation

Underwater imaging has long been focused on dehazing and color correction to address severe degradation in the water medium. In this paper, we propose a learning-based image restoration method that uses Generative Adversarial Networks (GAN). For network generality and learning flexibility, we constituted unpaired image translation frameworks into image restoration. The proposed method utilizes multiple cyclic consistency losses that capture image characteristics and details of underwater images. To prepare unpaired images of clean and degraded scenes, we collected images from Flickr and filtered out false images using image characteristics. For validation, we extensively evaluated the proposed network on simulated and real underwater hazy images. Also, we tested our method on conventional computer vision algorithms, such as the level of edges and feature matching results.

Endoscopic image feature matching via motion consensus and global bilateral regression

Background and objectiveFeature matching of endoscopic images is of crucial importance in many clinical applications, such as object tracking and surface reconstruction. However, with the presence of low texture, specular reflections and deformations, the feature matching methods of natural scene are facing great challenges in minimally invasive surgery (MIS) scenarios. We propose a novel motion consensus-based method for endoscopic image feature matching to address these problems. MethodsOur method starts by correcting the radial distortion with the spherical projection model and removing the specular reflection regions with an adaptive detection method, which helps to eliminate the image distortion and to reduce the quantity of outliers. We solve the matching problem with a two-stage strategy that progressively estimates a consensus of inliers; the result is a precisely smoothed motion field. First, we construct a spatial motion field from candidate feature matches and estimate its maximum posterior with expectation maximization algorithm, which is computationally efficient and able to obtain smoothed motion field quickly. Second, we extend the smoothed motion field to the affine domain and refine it with bilateral regression to preserve locally subtle motions. The true matches can be identified by checking the difference of feature motion against the estimated field. ResultsEvaluations are implemented on two simulation datasets of deformation (218 images) and four different types of endoscopic datasets (1032 images). Our method is compared with three other state-of-the-art methods and achieves the best performance on affine transformation and nonrigid deformation simulations, with inlier ratio of 86.7% and 94.3%, sensitivity of 90.0% and 96.2%, precision of 88.2% and 93.9%, and F1-Score of 89.1% and 95.0%, respectively. On clinical datasets evaluations, the proposed method achieves an average reprojection error of 3.7 pixels and a consistent performance in multi-image correspondence of sequential images. Furthermore, we also present a surface reconstruction result from rhinoscopic images to validate the reliability of our method, which shows high-quality feature matching results. ConclusionsThe proposed motion consensus-based feature matching method is proved effective and robust for endoscopic images correspondence. This demonstrates its capability to generate reliable feature matches for surface reconstruction and other meaningful applications in MIS scenarios.

Real-Time Dense Reconstruction of Tissue Surface From Stereo Optical Video.

We propose an approach to reconstruct dense three-dimensional (3D) model of tissue surface from stereo optical videos in real-time, the basic idea of which is to first extract 3D information from video frames by using stereo matching, and then to mosaic the reconstructed 3D models. To handle the common low-texture regions on tissue surfaces, we propose effective post-processing steps for the local stereo matching method to enlarge the radius of constraint, which include outliers removal, hole filling, and smoothing. Since the tissue models obtained by stereo matching are limited to the field of view of the imaging modality, we propose a model mosaicking method by using a novel feature-based simultaneously localization and mapping (SLAM) method to align the models. Low-texture regions and the varying illumination condition may lead to a large percentage of feature matching outliers. To solve this problem, we propose several algorithms to improve the robustness of the SLAM, which mainly include 1) a histogram voting-based method to roughly select possible inliers from the feature matching results; 2) a novel 1-point RANSAC-based [Formula: see text] algorithm called as DynamicR1PP [Formula: see text] to track the camera motion; and 3) a GPU-based iterative closest points (ICP) and bundle adjustment (BA) method to refine the camera motion estimation results. Experimental results on ex- and in vivo data showed that the reconstructed 3D models have high-resolution texture with an accuracy error of less than 2 mm. Most algorithms are highly parallelized for GPU computation, and the average runtime for processing one key frame is 76.3 ms on stereo images with 960×540 resolution.

Matching Algorithm of Statistical Optimization Feature Based on Grid Method

The matching algorithm based on image feature points is widely used in image retrieval, target detection, identification and other image processing fields. Aiming at the problem that the feature matching algorithm has low recall rate, a statistical optimization feature based on grid of the normalized cross correlation function is proposed. The matching main direction difference and scale ratio are introduced to feature matching process, for comprehensively utilizing SIFT feature points' information, such as the main direction, scale and position constrains, to accelerate the solution of matching position constraint under the grid framework, which optimizes the feature matching results and improves the recall rate and comprehensive match performance. Firstly, the nearest neighbor matching feature points of each feature point in the original image are found in the target image, and the initial matching results are obtained. Secondly, the matching main direction difference is used to eliminate most mismatches of the initial matching. Thirdly, the matching images are meshed based on the matching scale ratio information, and the position information of the matching feature points distributed among the grids is gathered statistics. Finally, the normalized cross correlation function of each small grid in the original image is calculated to determine whether the matching in the small grid is correct, and the optimized feature matching results are obtained. The experimental results show that the matching accuracy of the new algorithm is similar to that of classical feature matching algorithms, meanwhile the matching recall rate is increased by more than 10%, and a better comprehensive matching performance is obtained.

The Integration of the Image Sensor with a 3-DOF Pneumatic Parallel Manipulator.

The study aims to integrate the image sensor for a three-axial pneumatic parallel manipulator which can pick and place objects automatically by the feature information of the image processed through the SURF algorithm. The SURF algorithm is adopted for defining and matching the features of a target object and an object database. In order to accurately mark the center of target and strengthen the feature matching results, the random sample and consensus method (RANSAC) is utilized. The ASUS Xtion Pro Live depth camera which can directly estimate the 3-D location of the target point is used in this study. A set of coordinate estimation calibrations is developed for enhancing the accuracy of target location estimation. This study also presents hand gesture recognition exploiting skin detection and noise elimination to determine the active finger count for input signals of the parallel manipulator. The end-effector of the parallel manipulator can be manipulated to the desired poses according to the measured finger count. Finally, the proposed methods are successfully to achieve the feature recognition and pick and place of the target object.

An efficient feature descriptor based on synthetic basis functions and uniqueness matching strategy

Feature matching is an important step for many computer vision applications. This paper introduces the development of a new feature descriptor, called SYnthetic BAsis (SYBA), for feature point description and matching. SYBA is built on the basis of the compressed sensing theory that uses synthetic basis functions to encode or reconstruct a signal. It is a compact and efficient binary descriptor that performs a number of similarity tests between a feature image region and a selected number of synthetic basis images and uses their similarity test results as the feature descriptors. SYBA is compared with four well-known binary descriptors using three benchmarking datasets as well as a newly created dataset that was designed specifically for a more thorough statistical T-test. SYBA is less computationally complex and produces better feature matching results than other binary descriptors. It is hardware-friendly and suitable for embedded vision applications.

Generation and Rendering of Interactive Ground Vegetation for Real-Time Testing and Validation of Computer Vision Algorithms

In recent years, the realistic rendering of natural outdoor sceneries has become more and more important in many application areas. Regarding the development process of new algorithms for computer vision applications, testing and evaluation in real outdoor environments is time-consuming or hardly applicable in many cases. As a result,artificial testing environments are used, which differ from real-world environments, especially regarding realistically reacting ground vegetation. Thus, developers try to simulate natural environments in real-time virtual reality applications, which are commonly known as Virtual Testbeds. Since the first basic usage of Virtual Testbeds several years ago, the image quality of recent virtual environments has almost reached a level close to photorealism even in real-time due to new rendering approaches and increasing processing power of current graphics hardware. Thus, Virtual Testbeds can recently be applied in application areas like computer vision, that strongly rely on realistic scene representations. In this article, we introduce a novel ground vegetation rendering approach that is capable of generating large sceneries with realistic appearance and excellent performance. Our approach features wind animation, as well as object-to-grass interaction and delivers realistically appearing grass and shrubs at all distances and from all viewing angles. This greatly improves immersion, as well as acceptance, especially in virtual training applications. Nevertheless, the rendered results also fulfill important requirements for the computer vision aspect, like plausible geometry representation of the vegetation as well as its consistence during the entire simulation. Feature detection and matching algorithms are applied to our approach in localization scenarios of mobile robots in natural outdoor environments. We will show how the quality of feature matching results is influenced by violating the static scene constraint in the Virtual Testbed, as observed in highly unstructured, real-world outdoor scenes with wind and object-to-vegetation interaction.