Mean Shift Method Research Articles

Advanced Lake Shoreline Extraction Approach by Integration of SAR Image and LIDAR Data

Noise and an abnormal distributed-image histogram is the main challenge of using SAR data. From this point of view, this study’s authors motivated the non-use of user-defined input parameters. To achieve this purpose, a fuzzy approach was proposed to extract shoreline from SENTINEL-1A data. The parameters in the processing of the SENTINEL-1A image were generated automatically with LIDAR-intensity-derived object-based segmentation results. The LIDAR-intensity image was segmented with the Mean-shift method. The corresponding result was used to estimate the input parameters for fuzzy clustering of the SENTINEL-1A image. Fuzzy segmentation was proposed, due to the expected large number of values regarding water and land classes except for the pixels along the shoreline. The memberships for land and water classes were separately computed. In the proposed approach, the results from LIDAR and SENTINEL-1A dataset are promising, with differences below 1 pixel (10 m) by evaluation with the used reference vector data.

Segmenting images with complex textures by using hybrid algorithm

Application of computer images processing technology to analyze materials microstructural images, particularly metallographic images, has received increasing attention. The metallographic images contain the mesoscopic information on structural relation and components of materials. Quantitative analysis of these images can help to correlate the materials structures to their performance and properties at various levels. There are two challengeable issues necessary to be resolved, i.e., automatic segmentation and classification of different microscopic structures in metallographic images. Since the metallographic images often contain complex textures, the segmentation of them is usually inaccurate with present methods. We propose a hybrid algorithm, which combines the Gaussian filter, the mean shift method, the FloodFill, the improved flow-based difference-of-Gaussians, and the clustering to resolve the issues. The experiment results and the comparative results show that our method is effective to segment and classify the microstructural elements in metallographic images with complex textures.

Moving-Object Tracking Algorithm Based on PCA-SIFT and Optimization for Underground Coal Mines

In view of the complex and changeable environment in underground coal mines, an improved algorithm based on the principal component analysis-scale invariant feature transform (PCA-SIFT) and mean shift is proposed to address the issues for which existing tracking algorithms are not adequate; for example, when differentiating between moving targets and the background, the tracking in the case of moving objects (e.g., confusion between foreground and background) is not optimal. This results in poor resolution and the inability to deal with very dusty conditions, scale change, and rotation. The proposed feature target tracking model was developed using the scale invariance property of the PCA-SIFT feature-extraction algorithm. Finally, the mean-shift method was used to track moving objects. The experimental results showed that the optimized algorithm for tracking moving objects was significantly better and more robust than the existing algorithm.

Extraction of diamond grain topography from diamond tool surface using 3D surface measurement coupled with image analysis

Characterization of protruded diamond grains in a diamond tool is essential to explain the variation in its machining performance. This paper proposes a method for extraction of diamond grains and reconstruction of their topography from an electroplated diamond bit, using the topographic data and optical micrographs acquired with a confocal laser microscope. Otsu method, mean shift method and texture based method were selected for segmentation of the optical micrographs and their performance was evaluated. Experiment results show that texture based method can satisfactorily extract protruded diamond grains; its result was used to reconstruct diamond grain topography successfully.

Multiscale and Multifeature Segmentation of High-Spatial Resolution Remote Sensing Images Using Superpixels with Mutual Optimal Strategy

High spatial resolution (HSR) image segmentation is considered to be a major challenge for object-oriented remote sensing applications that have been extensively studied in the past. In this paper, we propose a fast and efficient framework for multiscale and multifeatured hierarchical image segmentation (MMHS). First, the HSR image pixels were clustered into a small number of superpixels using a simple linear iterative clustering algorithm (SLIC) on modern graphic processing units (GPUs), and then a region adjacency graph (RAG) and nearest neighbors graph (NNG) were constructed based on adjacent superpixels. At the same time, the RAG and NNG successfully integrated spectral information, texture information, and structural information from a small number of superpixels to enhance its expressiveness. Finally, a multiscale hierarchical grouping algorithm was implemented to merge these superpixels using local-mutual best region merging (LMM). We compared the experiments with three state-of-the-art segmentation algorithms, i.e., the watershed transform segmentation (WTS) method, the mean shift (MS) method, the multiresolution segmentation (MRS) method integrated in commercial software, eCognition9, on New York HSR image datasets, and the ISPRS Potsdam dataset. Computationally, our algorithm was dozens of times faster than the others, and it also had the best segmentation effect through visual assessment. The supervised and unsupervised evaluation results further proved the superiority of the MMHS algorithm.

Self-adaptive segmentation of satellite images based on a weighted aggregation approach

Image segmentation is a decisive process in object-based image analysis, while the uncertainty of segmentation scales can significantly influence the results. To resolve this issue, this study proposes a Self-Adaptive Segmentation (SAS) method which bridges the gap between the inherent scale and segmentation scale of each object. Firstly, SAS is defined as a variable-scale segmentation approach, aiming at generating local optimum results. It is then implemented based on Segmentation by Weighted Aggregation (SWA) method and optimized by selfhood scale estimation technique. Secondly, two images derived from WorldView-2 (an urban area) and Landsat-8 (a farmland area) are employed to verify the effectiveness and adaptability of this method. Thirdly, it is further compared with SWA and mean shift method by reference to two evaluations, including an unsupervised evaluation, i.e., Global Score (GS), and a supervised one named Geometry-based comparison. The experimental results indicate the SAS method presented in this study is effective in solving the scale issues in multiscale segmentations. It shows higher accuracy than SWA through assigning an appropriate segmentation scale for each object. Moreover, the segmentation results achieved by SAS method in both study areas are more accurate than those by mean shift method, which demonstrates the stable performance of SAS method across diverse scenes and data sets. This advantage is more obvious in farmland area than in urban scene, according to the accuracy assessment results.

A Target Tracking Algorithm Based on Mean Shift and Fast Template Matching

This paper proposes a target tracking algorithm based on mean shift and template matching. The algorithm is divided into three stages:prediction, template matching, target positioning, and template updating. In the prediction stage, combined with the target position obtained from the previous frame tracking, the target position is predicted using the mean shift method, and the template matching search gate is defined with the predicted position as the center and the corresponding size as the coverage area. At the template matching stage, using fast template matching algorithm, the target template and search gate are quickly matched from coarse to fine, and the matching degree between matching result and target template is calculated. If the matching degree is greater than the given threshold, the fast template matching will be performed and the result will be used as the tracking result of the current frame image. Otherwise, the target position predicted by the mean shift algorithm is used as the tracking results of the current frame image. Finally, the template updating process is controlled by the tracking results of the current frame to update the target template, and the stable tracking of the target is finally completed. At the same time, the algorithm improves the robust of tracking by combining the advantages of color and edge features to the insensitivity of rotation and deformation. The method has fast calculation speed and high accuracy, it can meet real-time requirements.

Depth sensor based human detection for indoor surveillance

Human detection is a popular topic and difficult problem in surveillance. This paper presents a research on human detection in complex indoor space utilizing a depth sensor. In recent years, target detection methods based on RGB-D data mainly include background learning, and feature detection operator. The former method depends on the initial background knowledge obtained from the first couple of frames in the video, and the length of frames decides detection quality. The latter method is more time intensive, and insufficient training samples will influence the detection result. Thus, in this paper we analyze the complex scene features thoroughly and integrate the color and depth information, proposing a RGBD+ViBe foreground extraction method. Based on the extraction outcome of the foreground, this research utilizes the 3D Mean-Shift method combined with depth constraints to handle multi-person partial occlusion problems. The experiment results indicate that the proposed RGBD+ViBe method outperforms the methods which only consider color or depth information, as well as the RGBD+MoG method. Furthermore, the proposed 3D Mean-Shift method achieves nearly 90% accuracy in multi-person detection result, and the false rate is merely 5%; while the accuracy of HOG, HOD and Comb-HOD methods are less than 75% and the false rate is around 10%.

Exudates detection in fundus images using mean-shift segmentation and adaptive thresholding

Diabetic retinopathy (DR) affects changes to retinal blood vessels that can cause them to bleed or leak fluid and distorting vision. An early detection of exudates is a prerequisite for detecting and grading severe retinal lesions, like DR. This paper presents an automated method for detection of the exudates in digital fundus images. Our approach can be divided into four steps: shifting colour correction, Optic disc (OD) elimination, exudates segmentation and separation of exudates from background. In order to correct non-uniform illumination, we adopted the grey world method. Then, we must extract the OD prior to the process because it appears with similar colour, intensity and contrast to exudates. Next, to segment the exudates, we applied the mean-shift method. Finally, we used the maximum entropy thresholding to separate the exudates from background. The proposed method is tested on DIARETDB0 and DIARETDB1. Comparing to other recent methods available in the literature, our proposed approach obtains better exudate detection results in terms of sensitivity, specificity and accuracy.

Algorithm based on regional separation for automatic grain boundary extraction using improved mean shift method

Metallographic microscopy shows that the vast majority of metal materials are composed of many small grains; the grain size of a metal is important for determining the tensile strength, toughness, plasticity, and other mechanical properties. In order to quantitatively evaluate grain size in metals, grain boundaries must be identified in metallographic images. Based on the phenomenon of grain boundary blurring or disconnection in metallographic images, this study develops an algorithm based on regional separation for automatically extracting grain boundaries by an improved mean shift method. Experimental observation shows that the grain boundaries obtained by the proposed algorithm are highly complete and accurate. This research has practical value because the proposed algorithm is suitable for grain boundary extraction from most metallographic images.

Saliency Detection Method Using Hypergraphs on Adaptive Multiscales

Saliency detection plays an important role in the fields of image processing and computer vision. We present an improved saliency detection method by means of hypergraphs on adaptive multi-scales (HAM). It first adjusts adaptively the ranges of pixel-values in ${R}, {G}, {B}$ channels in an input image and uses the three ranges to determine adaptive scales for the construction of hypergraphs. And then, the image is modeled as multiple hypergraphs on adaptive scales in which hyper-edges are clustered by means of the agglomerative mean-shift method. The HAM method can get more single-scale hypergraphs and thus has higher accuracy than the previous ones because hypergraphs are constructed on adaptive multi-scales instead of fixed scales. Extensive experiments on the published benchmark datasets demonstrate that the HAM method has improved the performance of detecting salient objects in input images, especially for the images with narrow ranges of pixel-values.

Fabrication and evaluation of regenerated cellulose/nanoparticle fibers from lignocellulosic biomass

This paper presents a new approach for converting lignocelluloses biomass into regenerated cellulose fibers. A major research objective is to examine the effectiveness of nanoparticles as cellulose fillers on improving thermal and mechanical properties of the regenerated cellulose fibers. Pure wood and bagasse cellulose solutions and cellulose/nanoparticle composite solutions were prepared using N-methylmorpholine-N-oxide (NMMO) solvent, and were spun into filament fiber through an extruding device. The thermal analysis methods of TGA and DSC were used for characterizing the solution properties. The regenerated bagasse cellulose fiber was also compared with regenerated wood cellulose fiber and nanoparticle/cellulose fibers in terms of mechanical strength and elongation. Distribution of nanoparticles in the cellulose matrix was evaluated using a method of image analysis for fiber photomicrographs. Tested results indicated that the tensile strength of the regenerated bagasse cellulose fiber was considerably lower than that of the regenerated wood cellulose fiber but the elongation was consistent between these two fibers. The research also revealed that adding a small amount of nanoparticles to fill in the wood cellulose matrix resulted in an increase of the cellulose fiber tensile strength and modulus by 14% and 6% respectively, and a decrease of cellulose glass transition temperature. The nanoparticle distribution in the cellulose matrix was correctly detected by a computing image technique based on a mean shift method.

A Novel Method for Person Tracking Based K-NN : Comparison with Sift and Mean Shift Method

A Novel Method for Person Tracking Based K-NN : Comparison with Sift and Mean Shift Method

Landsat-Based Land Cover Change in the Beijing-Tianjin-Tangshan Urban Agglomeration in 1990, 2000 and 2010

Rapid urbanization dramatically changes the local environment. A hybrid classification method is designed and applied to multi-temporal Landsat images and ancillary data to obtain land cover change datasets. A support vector machine (SVM) classifier is used to classify multi-temporal Landsat Enhanced Thematic Mapper Plus (ETM+) images that were collected in 2000 at the pixel level. These images are also segmented with the mean shift method. The impervious surface is refined based on a combination of the segmented objects and the SVM classification results. The changed areas in 1990 and 2010 are determined by comparing the Thematic Mapper (TM) and ETM+ images via the re-weighted multivariate alteration detection transformation method. The TM images that were masked as changed areas in 1990 and 2000 are input into the SVM classifier. Land cover maps for 1990 and 2010 are produced by combining the unchanged area in 2000 with the new classes of the changed areas in 1990 and 2010. Land cover change has continuously accelerated since 1990. Remarkably, arable land decreased, while the impervious surface area significantly increased.

Independent Object Tracking from Video using the Contour Information in HSV Color Space

Background/Objectives: In this paper, a tracking method that is based on the colors and contours is proposed to obtain results that are superior to those of the traditional methods for the tracking of the objects in video. Methods/Statistical Analysis: The proposed method requires an initial conversion of the existing RGB-color-based video in the HSV color space, and a back-projection histogram in the HSV color space is used to obtain the color information of the object region that is to be tracked. The contour information is also obtained to track the target based on the Canny algorithm, whereby the moving object in the video is tracked in real time on the basis of the color and contour information. Findings: An experiment was performed wherein the proposed method was compared with the Mean-shift tracking method, the Camshaft tracking method, the Kalman-filter tracking method, and the Particle-filter tracking method in terms of the objective of this paper. The comparison showed that the proposed method is more accurate regarding the tracking of a single object in video. Improvements/Applications: This improved tracking method will be utilized in object tracking, CCTV for car tracking and HCI.

Defining Crowd Movement as Parabola and Classifying These Definitions

Smart surveillance systems developed in recent years have made enormous contributions to providing safety and management of crowds. The aim of this study is to observe and try to understand how crowd movements presented in a video sequence show behaviour. For this end, the motion data at pixel level among the consecutive frames is obtained using optical flow initially. Then, this motion data is associated using the particle advection method and stable as well as moving areas in the image are obtained. After, the moving areas clustered using Mean-Shift method are described and classified as parabola, in addition to the studies in the literature. At the end of the study, the method developed was tested over UCF as well as Pets2009 datasets and the results are presented.

Reweighted mass center based object-oriented sparse subspace clustering for hyperspectral images

Considering the inevitable obstacles faced by the pixel-based clustering methods, such as salt-and-pepper noise, high computational complexity, and the lack of spatial information, a reweighted mass center based object-oriented sparse subspace clustering (RMC-OOSSC) algorithm for hyperspectral images (HSIs) is proposed. First, the mean-shift segmentation method is utilized to oversegment the HSI to obtain meaningful objects. Second, a distance reweighted mass center learning model is presented to extract the representative and discriminative features for each object. Third, assuming that all the objects are sampled from a union of subspaces, it is natural to apply the SSC algorithm to the HSI. Faced with the high correlation among the hyperspectral objects, a weighting scheme is adopted to ensure that the highly correlated objects are preferred in the procedure of sparse representation, to reduce the representation errors. Two widely used hyperspectral datasets were utilized to test the performance of the proposed RMC-OOSSC algorithm, obtaining high clustering accuracies (overall accuracy) of 71.98% and 89.57%, respectively. The experimental results show that the proposed method clearly improves the clustering performance with respect to the other state-of-the-art clustering methods, and it significantly reduces the computational time.

Intelligent Video Surveillance System Based on Moving Object Detection and Tracking

Aiming at the problem that traditional video surveillance system is not intelligent, a new idea of intelligent video surveillance system is proposed. Analysis of the commonly used moving object detection and tracking methods, using three frame difference and Mean Shift method for automatic detection and tracking of moving objects in video surveillance. A suspicious event rules corresponding to the definition of all kinds of security threats, application of moving object detection and tracking method, the realization of the monitoring video system of automatic alarm function, can improve the level of intelligent video surveillance system.

Extended social force model‐based mean shift for pedestrian tracking under obstacle avoidance

It has been shown that mean shift tracking algorithm can achieve excellent results in pedestrian tracking task. It empirically estimates the target position of current frame by locating the maximum of a density function from the local neighborhood of the target position of previous frame. However, this method only considers its past trajectory without considering the influence of pedestrian environment when applying to pedestrian tracking. In practical, pedestrians always keep a safe distance away from obstacles when programming their paths. To address the issue of obstacle avoidance, this paper proposes a novel extended social force model‐based mean shift tracking algorithm in which pedestrian environment is full taken in consideration. Firstly, an extended social force model is presented to quantify the interaction between pedestrian and obstacle by means of force. Furthermore, directional weights and speed weights are introduced to adjust the strength of the force in terms of the difference of individual perspectives and relative velocities. Finally, the initial target position is predicted by Newton's laws of motion and then the Mean Shift method is integrated to track target position. Experiment results show that this algorithm achieves an encouraging performance when obstacles exist.

Efficient Combination of Texture and Color Features in a New Spectral Clustering Method for PolSAR Image Segmentation

In this paper, an efficient combination of texture and color features is proposed for polarimetric synthetic aperture radar (PolSAR) image segmentation. In proposed method, the PolSAR image is first segmented using the mean-shift method. Then, in each segment obtained from the mean shift, the first moment of color for the three color components of L* a* b* color space is obtained. On the other hand, a texture feature vector for each pixel of the image is formed corresponding to the texture edge energy at different directions with Gabor filter. Then, a new spectral clustering method is used to combine texture and color features. The results show that the proposed method is effective in PolSAR image segmentation.