Feature Point Clustering Research Articles

Research on the intelligent detection technology of multi-objective coarse aggregates

In order to achieve the screening of multi-objective coarse aggregate particles and the detection of needle-like particles, this paper targets the limestone coarse aggregate feature points obtained by Harris corner detection as the research object, using unsupervised clustering technology to intelligently cluster the corner coordinates. To achieve intelligent separation of coarse aggregates and screening of needle-like particles. At the same time, the Harris corners are sensitive to scale, the multi-scale corner information is fitted and the sieving critical value is regressed, and the needle flake determination threshold is integrated. Finally, the intelligent screening system of multi-objective coarse aggregate was integrated, and the correct recognition rate was counted. Studies have shown that the number of feature points of coarse aggregates with different particle sizes has a stable change at different resolutions. Intelligent clustering of feature points of coarse aggregates can replace image segmentation. There is a strong correlation between coarse aggregate texture information and particle size.

Visual detection of rice rows based on Bayesian decision theory and robust regression least squares method.

Paddy field management is complicated and labor intensive. Correct row detection is important to automatically track rice rows. In this study, a novel method was proposed for accurate rice row recognition in paddy field transplanted by machine before the disappearance of row information. Firstly, Bayesian decision theory based on the minimum error was used to classify the period of collected images into three periods (T1: 0-7 d; T2: 7-28 d; T3: 28-45 d), and resulting in the correct recognition rate was 97.03%. Moreover, secondary clustering of feature points was proposed, which can solve some problems such as row breaking and tilting. Then, the robust regression least squares method (RRLSM) for linear fitting was proposed to fit rice rows to effectively eliminate interference by outliers. Finally, a credibility analysis of connected region markers was proposed to evaluate the accuracy of fitting lines. When the threshold of credibility was set at 40%, the correct recognition rate of fitting lines was 96.32%. The result showed that the method can effectively solve the problems caused by the presence of duckweed, high-density inter-row weeds, broken rows, tilting (±60°), wind and overlap. Keywords: rice rows detection, Bayesian decision theory, clustering, RRLSM, credibility analysis, automatic tracking DOI: 10.25165/j.ijabe.20211401.5910 Citation: He J, Zang Y, Luo X W, Zhao R M, He J, Jiao J K. Visual detection of rice rows based on Bayesian decision theory and robust regression least squares method. Int J Agric & Biol Eng, 2021; 14(1): 199–206.

Estimation of Flat Object Deformation Using RGB-D Sensor for Robot Reproduction.

This paper introduces a system that can estimate the deformation process of a deformed flat object (folded plane) and generate the input data for a robot with human-like dexterous hands and fingers to reproduce the same deformation of another similar object. The system is based on processing RGB data and depth data with three core techniques: a weighted graph clustering method for non-rigid point matching and clustering; a refined region growing method for plane detection on depth data based on an offset error defined by ourselves; and a novel sliding checking model to check the bending line and adjacent relationship between each pair of planes. Through some evaluation experiments, we show the improvement of the core techniques to conventional studies. By applying our approach to different deformed papers, the performance of the entire system is confirmed to have around 1.59 degrees of average angular error, which is similar to the smallest angular discrimination of human eyes. As a result, for the deformation of the flat object caused by folding, if our system can get at least one feature point cluster on each plane, it can get spatial information of each bending line and each plane with acceptable accuracy. The subject of this paper is a folded plane, but we will develop it into a robotic reproduction of general object deformation.

Curve Skeleton Extraction From 3D Point Clouds Through Hybrid Feature Point Shifting and Clustering

AbstractCurve skeleton is an important shape descriptor with many potential applications in computer graphics, visualization and machine intelligence. We present a curve skeleton expression based on the set of the cross‐section centroids from a point cloud model and propose a corresponding extraction approach. We first provide the substitution of a distance field for a 3D point cloud model, and then combine it with curvatures to capture hybrid feature points. By introducing relevant facets and points, we shift these hybrid feature points along the skeleton‐guided normal directions to approach local centroids, simplify them through a tensor‐based spectral clustering and finally connect them to form a primary connected curve skeleton. Furthermore, we refine the primary skeleton through pruning, trimming and smoothing. We compared our results with several state‐of‐the‐art algorithms including the rotational symmetry axis (ROSA) and L1‐medial methods for incomplete point cloud data to evaluate the effectiveness and accuracy of our method.

A Cerebrovascular Image Segmentation Method Based on Geometrical Feature Point Clustering and Local Threshold

A Cerebrovascular Image Segmentation Method Based on Geometrical Feature Point Clustering and Local Threshold

Automated Counting and Tracking of Vehicles

A robust traffic surveillance system is crucial in improving the control and management of traffic systems. Vehicle flow processing primarily involves counting and tracking vehicles; however, due to complex situations such as brightness changes and vehicle partial occlusions, traditional image segmentation methods are unable to segment and count vehicles correctly. This paper presents a novel framework for vision-based vehicle counting and tracking, which consists of four main procedures: foreground detection, feature extraction, feature analysis, and vehicles counting/tracking. Foreground detection intends to generate regions of interest in an image, which are used to produce significant feature points. Vehicles counting and tracking are achieved by analyzing clusters of feature points. As for testing on recorded traffic videos, the proposed framework is verified to be able to separate occluded vehicles and count the number of vehicles accurately and efficiently. By comparing with other methods, we observe that the proposed framework achieves the highest occlusion segment rate and the counting accuracy.

Research on Mining Cloud Data Based on Correlation Dimension Feature

Large hierarchical cloud storage database has distributed of non-continuous massive data, the data has nonlinear characteristics of strong coupling, and using traditional methods for data mining, mining exist difficult problems. This paper proposes mining algorithm based a cloud non continuous layer data, and analyze the overall data mining model. The paper use fuzzy C means clustering algorithm to complete the semantic ontology feature point clustering beam based on semantic feature extraction and quantization encoding, to realize improved data mining algorithm. The experimental results show that the improved algorithm, the non-continuous mining level data have high precision, good performance, anti-interference ability strong, performance is superior to the traditional method.

Facial Animation of Life - Like Avatar based on Feature Point Cluster

Facial Animation of Life - Like Avatar based on Feature Point Cluster

Cloud tracking using clusters of feature points for accurate solar irradiance nowcasting

In this work, we propose a system to track the clouds and predict relevant events based on all-sky images. To deal with the nature of variable appearance of clouds, we use clusters of feature points to perform tracking. We propose an enhanced clustering algorithm that does not require prior knowledge of number of clusters. The proposed clustering algorithm can successfully separate feature points into groups with reasonable sizes and ranges. In the tracking process, merging and splitting of clouds are handled via checking matched pairs of feature points among different clusters. Afterwards, the tracking information is utilized to predict if the sun will be covered or obscured by clouds within the prediction horizon. Features are extracted from the tracked feature points and a Markov chain model is designed to perform ramp-down event prediction. The obscuration and ramp-down events have an important impact on solar irradiance. The experiments have shown that the proposed system can substantially enhance the accuracy of solar irradiance nowcasting on a challenging dataset.

Research on New Multi-Feature Large-Scale Image Retrieval Algorithm based on Semantic Parsing and Modified Kernel Clustering Method

Because of the feature points can describe the local characteristics of the image in a reasonable manner, effective use of feature point of content based image retrieval become the current hot issues in the field of computer vision. Aiming at this problem, we put forward a kind of combination clustering based on feature points, a new method of image retrieval. The method includes the combination of feature point clustering algorithm and based on the algorithm of local color histogram construction strategy. With the existing and local color histogram retrieval method based on feature points, compared to the method can effectively solve the current method of feature point location information and feature point center relying too much on the problem. Subjectivity and as a result of the manual annotation image accuracy, the traditional image retrieval methods cannot meet the needs of the user. Multidimensional indexing technology is only from the perspective of how to improve the indexing algorithm to adapt to the large-scale database to consider a problem, in content-based image retrieval. Our research combines the advantages of the semantic analysis and kernel clustering which will enhance the performance of the traditional image retrieval methods and strengthen the feasibility of the algorithm.

Counting crowd flow based on feature points

A counting approach for crowd flow based on feature points is proposed. The objective is to obtain the characteristics of the crowd flow in a scene, including the crowd orientation and numeric count. For the feature point detection, a three-frame difference algorithm is used to obtain a foreground containing only the moving objects. Therefore, after the SURF feature point detection, only the feature points of the foreground are retained for further processing. This greatly reduces the time complexity of the SURF algorithm. For feature point clustering, we present an improved DBSCAN clustering algorithm in which the non-motion feature points are further eliminated and only the remaining feature points are clustered. For the calculation of the crowd flow orientation, the feature points are tracked based on a local Lucas–Kanade optical flow with Hessian matrix algorithm. In the crowd flow number counting, the crowd eigenvectors are constructed based on the SURF feature points and are trained using a support vector regression machine. The experimental results show that the proposed crowd orientation and counting method are more robust and provide crowd flow statistics with higher accuracy than previous approaches.

Technique of Line-Carved Relief Editing Based on Mesh Feature Operations

A methodology of line-carved relief editing by mesh feature operations was proposed. Feature region of mesh model is firstly identified by curvature criterion. Secondly, profile lines are extracted by such feature operations as clustering, eliminating, retrieving and thinning of feature points. Thirdly, profile lines are optimized by trimming and smoothing operations. Finally, shape of carved curves could be edited by specifying the width and depth of profile lines, and a satisfactory line-carved relief is achieved by nonlinear compression.

Content‐based image authentication by feature point clustering and matching

ABSTRACTDigital multimedia makes fabricating and copying much easier than ever before. Therefore, it demands efficient and automatic techniques to identify and verify the content of digital multimedia. Image authentication is such a technique to automatically identify whether the query image is a fabrication or a simple copy of the original one. In this paper, we propose a perceptual image authentication technique based on clustering and matching of feature points of images. Feature points are first extracted from images with the k‐largest local total variations and clustered using Fuzzy C‐means clustering algorithm. Then, feature points in the query image and the anchor image are matched into pairs in zigzag ordering along the diagonals of the images cluster by cluster. In the mean time, the outliers of feature points are removed. Then, the system decisions about the authenticity of images are determined by the majority vote of whether three types of distance between matched feature point pairs are larger than their respective thresholds. The three types of distance include the following: (i) histogram‐weighted distance, which is proposed in this paper; (ii) the normalized Euclidean distance; and (iii) the Hausdorff distance. The geometric transform between the query image and the anchor image is estimated, and the query image is registered. The possible tampered image blocks are detected, and the percentage of the tampered area is roughly estimated. The experimental results show the effectiveness and robustness of the proposed image authentication system. Copyright © 2011 John Wiley & Sons, Ltd.

A CQPOCS Algorithm for Fast Feature Point Matching

A clustering and quadratic programming based projection onto convex sets(CQPOCS)algorithm for fast feature point matching is presented in this paper．Via feature point clustering,the problem of matching model point set and taget point set is converted in to the problem of matching corresponding clusters,thus reducing the computational cost．Then,quatratic programming based POCS algorithm is used to solve the cluster matching problem without incurring the successive POCS algorithm’s accumulating deviation due to successive projections onto row convex sets and column convex sets．Simulation results show that our CQPOCS algorithm has satisfactory matching accuracy and computational safety．

A randomized approach with geometric constraints to fingerprint verification

In this paper, a fuzzy bipartite weighted graph model is proposed to solve fingerprint verification problem. A fingerprint image is preprocessed first to form clusters of feature points, which are called feature point clusters. Twenty-four attributes are extracted for each feature point cluster. The attributes are characterized by fuzzy values. Attributes of an input image to be verified are considered as the set of left nodes in a fuzzy bipartite weighted graph, and the attributes of claimed template fingerprint image are considered as the set of right nodes in the graph. The fingerprint verification problem is thus converted into a fuzzy bipartite weighted graph matching problem. A matching algorithm is proposed for the fuzzy bipartite weighted graph model to find an optimal matching with a goodness score. Experimental results reveal the feasibility of the proposed approach in fingerprint verification.

Extraction of characters from form documents by feature point clustering

Among various kinds of documents, forms are the important types. The automatic processing of form documents is a problem which is essential to the advancement of office automation. The extraction of characters from form documents is a prerequisite for optical character recognition. In this paper, we will present a clustering-based technique for extracting characters from form documents. In this method, we treat the character extraction process as a pattern clustering problem. The feasibility of the novel method is demonstrated through experimenting various kinds of forms. Experimental results reveal the feasibility of the novel method.