Showing papers on "Corner detection published in 2022"
••
TL;DR: In this article , a fast corner detector with a simple architecture and high parallel computing characteristics is proposed, which can achieve or even exceed the detection accuracy of multi-scale analysis based detectors and clearly higher than other types of corner detectors.
Abstract: Multi-scale analysis based corner detection algorithms yield impressive performance, however, they are not efficient and not suitable for real-time computer vision tasks. The classical corner detection algorithms including FAST and Harris are computationally efficient, but their detection accuracy and repeatability are insufficient. This paper describes a novel fast corner detector with a simple architecture and high parallel computing characteristics. In order to simplify the corner detection architecture and improve its parallel computing performance, a new type of filter is proposed that can enhance corners and suppress edges and noise simultaneously. Then a novel efficient corner detector is proposed, which can be adapted to achieve real-time detection in hardware. Experimental results show that, with a very low computational cost and simple architecture, the proposed detector can achieve or even exceed the detection accuracy of multi-scale analysis based detectors. Its repeatability is similar to multi-scale analysis based detectors and clearly higher than other types of corner detectors. Therefore, it is potentially useful as an efficient corner detector for computer vision applications especially for portable real-time tasks.
9 citations
••
TL;DR: Zhang et al. as mentioned in this paper designed an effective corner feature representations network based on the characteristics of corners and proposed a novel loss function to minimize the localization error between the corner positions of the original image block and the transformed image blocks.
Abstract: Interest points (corners and blobs) play an important role in computer vision tasks such as image matching, image retrieval, and 3D reconstruction. Existing deep learning based interest point detection methods mainly focus on the interest point detection with high repeatability under image affine transformations while neglecting the importance of the characteristics of interest points. This will affect the detection and localization accuracy of interest points. In this paper, we design an effective corner feature representations network based on the characteristics of corners. The designed network has the ability to effectively learn corner feature information from images. A novel loss function is proposed to minimize the localization error between the corner positions of the original image block and the transformed image blocks. Furthermore, a novel corner detection architecture is proposed. The criteria on detection accuracy, localization accuracy, average repeatability, region repeatability, and image matching score are used to evaluate the proposed method against fourteen state-of-the-art methods. The experimental results show that the proposed performs significantly better than the state-of-the-arts.
6 citations
••
TL;DR: Zhang et al. as mentioned in this paper proposed a one-stage anchor-free network based on searching four corner points of an object, which can yield an arbitrary quadrilateral to fit objects with different shapes and orientations.
Abstract: Oriented object detection in remote sensing images has drawn great attention since it can provide more accurate bounding boxes. We propose a one-stage anchor-free network based on searching four corner points of an object, which can yield an arbitrary quadrilateral to fit objects with different shapes and orientations. We detect the corners by combining two strategies, where one regresses to the relative corner positions with respect to their corresponding center and the other directly detects the absolute corner positions from the corner heatmaps. By defining a candidate corner region based on the regressed results, we check whether corner points from the corner heatmaps are included in the region. If so, the closest one relative to the regressed corner is selected as the final position; otherwise, the regressed corner position is utilized. Experiments were conducted on two aerial remote sensing datasets, and the results demonstrated that the proposed method achieves superior performance to both the anchor-based and anchor-free methods.
4 citations
••
TL;DR: A FAST-Harris fusion corner detection algorithm is proposed to improve the shortcomings of the Harris algorithm, such as the low detection accuracy and low positioning accuracy, and a corner detection fusion model is established.
Abstract: Corner detection is a common method to obtain image features, and the detection effect influences the performance of matching and tracking directly. A FAST-Harris fusion corner detection algorithm is proposed to improve the shortcomings of the Harris algorithm, such as the low detection accuracy and low positioning accuracy, and a corner detection fusion model is established. First, the detected target image is padded, and then the FAST algorithm is used with a 25% reduced contrast points to achieve fast capture roughly; in this way, a candidate corner set is obtained. Then, screening the candidate corner is set one by one by calculating the response function of the Harris with Scharr operator to achieve capture accurately. Finally, the real corners are obtained using SAD for nonmaximum suppression. The positioning error, error detection rate, robustness, and running time of corner detection are obtained by the PyCharm platform. Compared with Harris, the error detection rate and localization error of the algorithm are reduced by 16.89% and 42.04%, respectively. Compared with 8 popular corner detection algorithms, the error detection rate and localization error of the algorithm in this paper are the lowest, which are 24.60% and 1.42 pixels. The robust performance in lossy JPEG compression is the best, with 17.37% shorter running time than Harris algorithm. The method in this paper can be used in scenarios such as autonomous driving and image search services.
4 citations
••
TL;DR: Wang et al. as mentioned in this paper fit multisegment curves in a quadratic form to the edges in a checkerboard-like marker, and the exact corner positions were considered as the intersections of the corresponding curves with analytical solutions.
Abstract: Checkerboard-like markers are widely applied in visual localization applications, including SLAM, augmented reality, robot navigation, and 3-D scene reconstruction. Most corner detectors assume that the marker is attached to a flat surface, which severely limits the placement of the marker and thus reduces the scope of use for these applications. However, in some scenarios, it is not easy to find an ideal flat surface required by most corner detectors in a given scenario, in which case the marker can often only be fixed on a curved surface. Therefore, the accuracy of most corner detectors may be reduced. In this study, a novel method is proposed with subpixel accuracy to detect and locate the corners on a chessboard-like marker, which is either flat or curved. The proposed method fits multisegment curves in a quadratic form to the edges in a checkerboard-like marker. The exact corner positions are considered as the intersections of the corresponding curves with analytical solutions. The proposed method achieves the state-of-the-art performance through experiments, including synthetic corner localization test, real-world stereo vision triangulation experiment, and pose estimation on a curved object, demonstrating the superiority of our method.
4 citations
••
TL;DR: Wang et al. as mentioned in this paper developed a novel measure for contour-based corner finding algorithm by using the multi-scale tangent-to-point distance (MSTPD) technique.
Abstract: Corner is an import type of feature point of image and has been widely used in image analysis and vision tasks. To enhance the robustness to contour noise and geometrical transformations and meanwhile improve the localization accuracy, we develop a novel measure for contour-based corner finding algorithm by using the multi-scale tangent-to-point distance (MSTPD) technique. First, image contour extraction is conducted; second, the curvature of the extracted contour is computed with the tangent-to-point distance technique under different scales. By introducing relative distance, MSTPD is more robust to geometric transformations and by employing multi-scale technique MSTPD is also robust to contour noise and much accurate on localization. Experimental results show that MSTPD is a promising corner detection scheme compared with the other seven impressive corner detection methods based on two common evaluation criteria, that is, average repeatability and localization error.
3 citations
••
TL;DR: The experimental results show that the proposed road-free space extraction and obstacle detection method based on stereo vision has good robustness and real-time performance for obstacle detection in a complex traffic environment.
Abstract: For the task of obstacle detection in a complex traffic environment, this paper proposes a road-free space extraction and obstacle detection method based on stereo vision. The proposed method combines the advantages of the V-disparity image and the Stixel method. Firstly, the depth information and the V-disparity image are calculated according to the disparity image. Then, the free space on the road surface is calculated through the RANSAC algorithm and dynamic programming (DP) algorithm. Furthermore, a new V-disparity image and a new U-disparity image are calculated by the disparity image after removing the road surface information. Finally, the height and width of the obstacles on the road are extracted from the new V-disparity image and U-disparity image, respectively. The detection of obstacles is realized by the height and width information of obstacles. In order to verify the method, we adopted the object detection benchmarks and road detection benchmarks of the KITTI dataset for verification. In terms of the accuracy performance indicators quality, detection rate, detection accuracy, and effectiveness, the method in this paper reaches 0.820, 0.863, 0.941, and 0.900, respectively, and the time consumption is only 5.145 milliseconds. Compared with other obstacle detection methods, the detection accuracy and real-time performance in this paper are better. The experimental results show that the method has good robustness and real-time performance for obstacle detection in a complex traffic environment.
2 citations
••
2 citations
••
2 citations
••
TL;DR: In order to solve the problem of multisolution and ill-formedness of the 3D reconstruction method of a single image (purpose), the author proposes a microscope image segmentation algorithm based on the Harris multiscale corner detection.
Abstract: In order to solve the problem of multisolution and ill-formedness of the 3D reconstruction method of a single image (purpose), the author proposes a microscope image segmentation algorithm based on the Harris multiscale corner detection. Separating complex engineering images into several simple basic geometric shapes, rebuild them separately to avoid the ill-conditioned solution problem of directly recovering depth information. In order to improve the registration accuracy of the corner-based image registration algorithm, the idea of multiresolution analysis was introduced into the classic Harris corner detection, and a gray intensity variation formula based on the wavelet transform was constructed, and a scale transformation characteristic was obtained so that the improved Harris corner detection algorithm is invariant to rotation, translation, and scale. Experimental results show that after reconstruction, the error between the length of the object measured based on the point cloud data and the actual length of the object is small, and both remain within the error range of 3 mm. The experiment verifies the fast, accurate, and stable characteristics of the improved algorithm.
2 citations
••
TL;DR: In this article , a markerless image alignment method for pressure-sensitive paint measurement data replacing the time-consuming conventional alignment method in which the black markers are placed on the model and are detected manually.
Abstract: We propose a markerless image alignment method for pressure-sensitive paint measurement data replacing the time-consuming conventional alignment method in which the black markers are placed on the model and are detected manually. In the proposed method, feature points are detected by a boundary detection method, in which the PSP boundary is detected using the Moore-Neighbor tracing algorithm. The performance of the proposed method is compared with the conventional method based on black markers, the difference of Gaussian (DoG) detector, and the Hessian corner detector. The results by the proposed method and the DoG detector are equivalent to each other. On the other hand, the performances of the image alignment using the black marker and the Hessian corner detector are slightly worse compared with the DoG and the proposed method. The computational cost of the proposed method is half of that of the DoG method. The proposed method is a promising for the image alignment in the PSP application in the viewpoint of the alignment precision and computational cost.
••
TL;DR: In this paper , a camera calibration method based on the EDLines algorithm for the automatic detection of chessboard corners was proposed, where the features of the broken straight lines at the corners were then used to filter the straight lines and remove the background straight lines outside the chessboard.
Abstract: To improve the robustness and accuracy of the corner-detection algorithm, this paper proposes a camera-calibration method based on the EDLines algorithm for the automatic detection of chessboard corners. The EDLines algorithm is initially used to perform straight-line detection on the calibration image. The features of the broken straight lines at the corners are then used to filter the straight lines and remove the background straight lines outside the chessboard. The pixels in the rectangular area around the filtered straight line are sorted by the gray gradient. After using the sorted results to fit the straight line, the coordinates of the intersection of the straight lines are taken as the initial coordinates of the corners and perform subpixel optimization on them. Finally, the corner points are sorted by the conversion between pixel-coordinate systems. The camera exposure time changes and complex imaging-background experiments show that the algorithm has no missed detection and redundancy in corner detection. The average reprojection error is found to be less than 0.05 pixels, which can be used in actual calibration.
••
01 Apr 2022TL;DR: In this article , a fingertip detection method based on Freeman chain code analysis is proposed in order to meet the requirements of real-time fingertip positioning and detection in some scenes, which can detect the fingertip position quickly and accurately, and has good detection performance.
Abstract: In order to meet the requirements of real-time fingertip positioning and detection in some scenes, a fingertip detection method based on Freeman chain code analysis is proposed in this paper. Firstly, the image collected by the camera needs to be preprocessed. The edge information of the palm is obtained by median filtering, analyzing the skin color space, binarization and morphological processing. The Freeman chain code of the hand contour is obtained by contour tracking calculation. After the convex points of the chain code of the hand contour are repaired, the matching points are analyzed according to the chain code difference, so as to realize the function of fingertip detection. The test shows that the fingertip detection based on Freeman chain code can detect the fingertip position quickly and accurately, and has good detection performance.
••
18 Jul 2022TL;DR: Wang et al. as discussed by the authors proposed a novel grouping algorithm, termed as Soft-Grouping Non-Maximum Suppression (SG-NMS), which merges grouping with NMS into a whole to simplify the pipeline and lift the efficiency.
Abstract: The grouping process of corner-based detectors still faces two challenges: 1) Hard-grouping. If one of the paired corners is wrongly estimated, the grouping goes wrong. 2) Complex pipeline. Vanilla methods regard corner grouping as an additional stage, complicating the post-processing. To eliminate these issues, we propose a novel grouping algorithm, termed as Soft-Grouping Non-Maximum Suppression (SG-NMS), which merges grouping with NMS into a whole to simplify the pipeline and lift the efficiency. SG-NMS is flexible to match the varied number of detected corners. Accordingly, we propose a multi-corners context enhanced module, Corner Visual Reasoning (CVR), as a grouping helper. Equipped with the proposed SG-NMS, a new multi-corners detector, SGCDet (Soft Grouping Corner-based Detector), is proposed. Experiments show that the inference speed of SGCDet is more than two times faster than state-of-the-art corner-based models with much higher accuracy.
••
TL;DR: Wang et al. as discussed by the authors proposed a novel cornerity measure based on a dynamic region of support (RoS), with which an efficient corner detector is developed, which can deliver superior performance and exhibit higher robustness over the existing contour-based corner detector.
Abstract: Existing contour-based corner detectors generally identify corners from a contour curve by measuring the cornerity of each point (i.e., the confidence to be a corner) with a fixed-radius region of support (RoS), and thus could yield inferior performance due to low adaptivity to local structures of the input curve. To overcome the difficulty, a novel cornerity measure based on a dynamic RoS is proposed in this paper, with which an efficient corner detector is developed. For a given point on the curve, the dynamic RoS is constructed with two straight-line arms stretching towards both sides along the curve, under a pre-determined error tolerance imposed on the average perpendicular distance from the curve to each arm within its stretching range. Then, our cornerity model is established based on the lengths of the two arms and the angle between them, which is then exploited to evaluate whether the current point is a corner or not via a cornerity thresholding. Extensive experimental results show that the proposed corner detector can deliver superior performance and exhibit higher robustness over the existing state-of-the-arts.
••
TL;DR: In this article , a set of benchmark evaluation metrics are suggested, including five conventional ones: the precision, the recall, the arithmetic mean of precision and recall (APR), the F score, the localization error (Le), and a new one proposed in this work called the repeatability referenced to ground truth (RGT).
Abstract: Corners are an important kind of image feature and play a crucial role in solving various tasks. Over the past few decades, a great number of corner detectors have been proposed. However, there is no benchmark dataset with labeled ground-truth corners and unified metrics to evaluate their corner detection performance. In this paper, we build three benchmark datasets for corner detection. The first two consist of those binary and gray-value images that have been commonly used in previous corner detection studies. The third one contains a set of urban images, called the Urban-Corner dataset. For each test image in these three datasets, the ground-truth corners are manually labeled as objectively as possible with the assistance of a line segment detector. Then, a set of benchmark evaluation metrics is suggested, including five conventional ones: the precision, the recall, the arithmetic mean of precision and recall (APR), the F score, the localization error (Le), and a new one proposed in this work called the repeatability referenced to ground truth (RGT). Finally, a comprehensive evaluation of current state-of-the-art corner detectors is conducted.
••
TL;DR: LuvHarris as discussed by the authors employs the Harris algorithm for high accuracy but manages an improved event throughput, which is a necessity when using a high-resolution event-camera in real-time.
Abstract: There have been a number of corner detection methods proposed for event cameras in the last years, since event-driven computer vision has become more accessible. Current state-of-the-art have either unsatisfactory accuracy or real-time performance when considered for practical use, for example when a camera is randomly moved in an unconstrained environment. In this paper, we present yet another method to perform corner detection, dubbed look-up event-Harris (luvHarris), that employs the Harris algorithm for high accuracy but manages an improved event throughput. Our method has two major contributions, 1. a novel "threshold ordinal event-surface" that removes certain tuning parameters and is well suited for Harris operations, and 2. an implementation of the Harris algorithm such that the computational load per event is minimised and computational heavy convolutions are performed only 'as-fast-as-possible', i.e., only as computational resources are available. The result is a practical, real-time, and robust corner detector that runs more than 2.6× the speed of current state-of-the-art; a necessity when using a high-resolution event-camera in real-time. We explain the considerations taken for the approach, compare the algorithm to current state-of-the-art in terms of computational performance and detection accuracy, and discuss the validity of the proposed approach for event cameras.
••
08 Apr 2022TL;DR: In this article , an improved self-adapting conner detection algorithm is proposed in order to suppress the impact of image dithering caused by wind-induced vibration, the features of sequential frame should be extracted.
Abstract: In order to suppress the impact of image dithering caused by wind-induced vibration, the features of sequential frame should be extracted. The Harris corner detection algorithm has been widely used for feature extraction. In the field of security monitoring and control, the image and video usually have the characteristics of jumbo size, high pixel and low contrast, which are difficult to obtain the corners. For the problems above, an improved self-adapting conner detection algorithm is proposed in this paper. Firstly, some of the corners are selected according to the comparison results between target pixel and the pixels around. Secondly, the selected corners are classified by reference to adaptive threshold values. Finally, false and marginal corners can be reduced or eliminated so as to select the best-matching corners. The above improved algorithm is validated in the field of sea area security monitoring and control. Simulation results show the effectiveness and feasibility of the algorithm above.
••
08 Apr 2022TL;DR: Wang et al. as mentioned in this paper proposed a simple and fast corner detection method, which can effectively overcome the data redundancy problem in traditional corner detection algorithms, and the experimental results show that the algorithm can effectively extract the target corner, which is basically consistent with the manually calibrated target point.
Abstract: In vision-based alignment systems, it is very important to detect the corner positions in images. It is based on these positions that the systems can calculate the offset between the workpieces and achieve their assembly alignment using robots. A simple and fast corner detection method is proposed, which can effectively overcome the data redundancy problem in traditional corner detection algorithms. First, the reduced image is figured out by scale transformation and down-sampling, and the corresponding edge binary image is obtained by gradient analysis. Then, all the edge coordinates are detected by the designed edge extraction operator, and an initial corner point is obtained by fitting analysis of these edge points. Due to the error arising from the above transformation and fitting, the point is not the target one. Finally, the Hough transform is used to detect the local linear features near the point, and the target point is determined by the linear intersection analysis. The experimental results show that the algorithm can effectively extract the target corner, which is basically consistent with the manually calibrated target point.
••
TL;DR: In this article , a novel bolt loosening angle detection method based on binocular vision is presented. But, the method is prone to error scaused by the camera perspective.
Abstract: Bolt looseness detection is critical in preventing bolt connection failure. Compared to traditional sensor-based bolt looseness detection, image-based methods are low-cost and contactless and have thus become the highlight of research. However, current monocular vision-based detection methods are prone to error scaused by the camera perspective . In this paper, we present a novel bolt loosening angle detection method based on binocular vision. Key points on the bolt are detected and matched by SuperPoint Gauss network for 3D coordinates reconstruction and motion tracking. The bolt loosening angle is solved by fitting the rotation equation using random sample consensus. Experiments verify the proposed method performs well under different perspectives of camera and illumination conditions with an average error of 1.5°. Comparative test shows our method is superior to the monocular vision-based method in terms of accuracy when there is a large perspective angle. The proposed method is mark-free and robust to various working conditions, which makes it of great value for engineering application.
••
23 Sep 2022
TL;DR: Experimental results show that the improved algorithm improves the corner detection rate and reduces the number of redundant corner points, which is better than the multi-scale Harris algorithm and Harris-Laplace algorithm.
Abstract: De-noising, enhancing, and edge extraction of ocean engineering images can improve the utilization of information. After bilateral filtering pre-processed the image, a multi-scale spatial representation is introduced, and finally an approximate sparse model iterative learning strategy is used to group and detect the candidate corner points set to select the corner points at the smallest scale as the best corner points. Experimental results show that the improved algorithm improves the corner detection rate and reduces the number of redundant corner points, which is better than the multi-scale Harris algorithm and Harris-Laplace algorithm.
••
TL;DR: In this paper, the Hessian corner detector with wavefront coding was proposed to solve the problem of missing and redundant images in the long depth of field camera by making use of the distance characteristics between corners of the checkerboard and constraining the angle characteristics.
••
08 Dec 2022
TL;DR: Zhang et al. as mentioned in this paper proposed an image correction method based on corner point detection to better achieve image correction effect, where the contours of the binarized image are extracted firstly and closed contours are filled to avoid the independent contours from affecting the accuracy of region growth.
Abstract: To better achieve image correction effect, an image correction method based on corner point detection is proposed. In image preprocessing, firstly, image equalization is achieved based on the contrast limited adaptive histgram equalization to avoid the problems caused by illumination and suppress noise while maintaining details, and then adaptive threshold segmentation is performed using the OTSU to obtain the binarized image. In the corner point detection stage, the contours of the binarized image are extracted firstly and the closed contours are filled to avoid the independent contours from affecting the accuracy of region growth, then the center point of the image is used as the seed pixel for region growth, and finally the four corner points are calculated based on the linear contours of the growth region and Hough theory, where the accurate region growth result can avoid the influence of the background on the corner point detection. In the correction stage, the perspective matrix is calculated by the four corner points, and the image is corrected by the perspective transformation. The experiments show that the proposed method can accurately find the corner points of document images and achieve efficient correction.
••
••
01 Jan 2022
TL;DR: In this article , corner points are compacted inside a vehicle region which is considered as the initial requirement for an algorithm and densely packed corners are grouped, then the non-vehicle region is segmented.
Abstract: AbstractThis paper presents a unique pattern of vehicle detection with the help of fundamental and successive algorithms. The characteristics of the vehicle are the important parameters to identify vehicles. A good number of corner points is compacted inside a vehicle region which is considered as the initial requirement for an algorithm. The densely packed corner points are grouped. This grouping gives a hint of points which are associated with each vehicle and they play a key in detection of vehicles. Once the grouping is performed, the non-vehicle region is segmented. The corner points are tracked with a Lucas-Kanade algorithm in order to maintain the stability of corner points. The detection rate with the proposed method is 93.95%.KeywordsCorner pointsTrackingGrouping
••
TL;DR: The paper presents a comparative analysis of the corner detection methods of the Moravec algorithm, the Harris algorithm, and the Shi-Tomasi algorithm.
Abstract: This paper discusses implementation of a machine vision system for pattern recognition of design estimates. The main problem in the development of these systems is the choice of unique features that remain invariant to various kinds of transformations. Angular descriptors were chosen as a dominant feature. The paper presents a comparative analysis of the corner detection methods of the Moravec algorithm, the Harris algorithm, and the Shi-Tomasi algorithm. The authors have developed software in Python language that implements the operation of the Harris detector and the Shih-Tomasi detector. The recognition system is being tested for building 3D models in Blender
••
01 Jul 2022TL;DR: Li et al. as discussed by the authors proposed an improved Harris algorithm, which is inspired by the Contrast Limited Adaptive Histogram Equalization (CLAHE), after extracting a patch of the target image, the gray value of the patch is adjusted based on the cumulative distribution function (CDF).
Abstract: Corner points are commonly defined as the intersection of two edges, and the Harris algorithm, which performs corner point detection based on the grey value variation between a patch and its neighborhood, is commonly used in various computer vision tasks. For low-light images, Harris algorithm is affected because the details of the image become blurred by low contrast. This paper proposes an improved Harris algorithm, which is inspired by the Contrast Limited Adaptive Histogram Equalization (CLAHE). After extracting a patch of the target image, the gray value of the patch is adjusted based on the cumulative distribution function (CDF). As a result, the gray value of the patch becomes evenly distributed, and the variation of the gray value of the patch becomes sharper. The improved Harris algorithm has been compared with the original Harris algorithm on different images of low-light scenes. Experimental results show that the proposed algorithm can effectively detect corner points in low contrast regions, and the repeatability of corner points matching in the low-light regions is significantly improved.
••
12 Aug 2022
TL;DR: In this article , a review of edge, corner, and boundary detection methods is presented, and the importance of image prepossessing to minimise the noise is discussed as well.
Abstract: This is a review paper of traditional approaches for edge, corner, and boundary detection methods. There are many real-world applications of edge, corner, and boundary detection methods. For instance, in medical image analysis, edge detectors are used to extract the features from the given image. In modern innovations like autonomous vehicles, edge detection and segmentation are the most crucial things. If we want to detect motion or track video, corner detectors help. I tried to compare the results of detectors stage-wise wherever it is possible and also discussed the importance of image prepossessing to minimise the noise. Real-world images are used to validate detector performance and limitations.
••
TL;DR: In this article , the Hessian corner detector with wavefront coding was used to detect missing and redundant corners in the long depth-of-field (LDOF) camera, which can obtain the whole object information at one time within 3-13 m of the depth of field.