The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Occlusion is one important problem in single object tracking. However, conventional methods are not capable of making full use of the spatial information because of occlusion, which may lead to the drift. In this paper, we propose a robust patches-based tracking method via sparse representation, namely RPSR, which selects the unoccluded patches, and adaptively assigns larger contribution factors to...
Pedestrian flow estimation is a vital issue in video surveillance. Inspired by fluid mechanics we proposed to model the pedestrian flow as time-dependent fluid, and estimate the pedestrian flow using flux. Firstly, optical flow is used to construct the motion vector field. Then, we compute the inside and outside flux components within fixed areas to estimate the pedestrian flow in different direction...
In this paper, we study the problem of fine-grained image categorization, which is much more useful in real applications than basic image classification. Based on the most challenge dataset, CUB-200, we combine Efficient match kernel (EMK) with the weighted spatial pyramid to achieve state-of-art performance. Comparison with BoW, which can also be viewed as kernel matching approach, EMK digs the relations...
Face recognition and verification is still a challenging problem due to several issues such as pose, facial expression, occlusion, imaging conditions, rotation, size and orientation. This paper addresses the problem of recognizing human faces despite the presence in pose and size variation. To handle these problems, we mainly focus on block size definition. Instead of uniform block we thus propose...
In this paper, it is shown that Local Zernike Moments which is used in object and face recognition applications succesfully, can also used for face-pair matching problem. In this study, instead of using feature vectors produced by LZM directly, we focussed on reducing the dimensions of feature vectors and increasing the performance. In the light of experimental results, a new method called L2ML-YZM...
In this paper, a low computational load based region covariance descriptor (RCD) approach has been proposed. The proposed method is based on using fewer feature vectors on construction of RCD. Number of feature vectors is reduced by utilizing the random pixel decimation pattern. By making use of the proposed method, fast covariance descriptor based object tracking has been carried out. As can be seen...
Many modern computer vision systems combine high dimensional features and linear classifiers to achieve better classification accuracy. However, the excessively long features are often highly redundant; thus dramatically increases the system storage and computational load. This paper presents a novel feature selection algorithm, namely cardinal sparse partial least square algorithm, to address this...
We propose a novel framework for fast and robust video anomaly detection and localization in complicated crowd scenes. Images of each video are split into cells for extracting local motion features represented by optical flow. In the train videos, most background cells are subtracted by ViBe model. Feature vectors are extracted from each cell by integrating the value of optical flow in 8 different...
Gender classification of depth images is a challenging problem, most research work attempted to use shape information to solve this problem in the past literature. In this work, we propose a new fusion scheme for gender classification using both texture and shape features. A new ensemble scheme is advocated to combine texture and shape feature at the feature level. To evaluate the performance of our...
In this paper, we propose a new method to detect abnormal behavior in crowd video. The motion influence matrix is proposed to represent crowd behaviors. It is generated based on concept of human perception with block-level motion vectors which describe actual crowd movement. Furthermore, a generalized framework is developed to detect abnormal crowd behavior using motion influence matrix. The proposed...
In most action recognition tasks, the target videos are of different temporal length and an action may be repeated several times in one video. As a result, encoding the distribution of video feature points that are far away from each other or are detected from different occurrences of the action into the video representation is inappropriate. It is better that sub video shots containing different...
Typically, the saliency map of an image is usually inferred by only using the information within this image. While efficient, such single-image-based methods may fail to obtain reliable results, because the information within a single image may be insufficient for defining saliency. In this paper, we propose a novel idea of learning with labeled images and adopt a new paradigm called sample specific...
We propose an approach to improving the detection results of a generic offline trained detector on frames from a specific video. For two consecutive frames of a video with the object, deformable part model(DPM) detection is perform to get the original detections. Then respectively obtain the image patches corresponding to the detected root box and part boxes. Thirdly, extract scale invariant feature...
This paper proposes a novel framework to calibrate cameras and model a scene simultaneously using both self-calibration constraints and geometric information on parallelograms. The proposed method is a factorization-based approach solving the problem by decomposing a measurement matrix into camera and parallelogram parameters. Since the factorizationbased approach recovers all camera poses simultaneously,...
This paper proposes a fast method for image matching under MPEG-7 CDVS Framework. For fast matching, 16-dimensional binary vector was constructed to represent a 128-dimensional SIFT descriptor and to present a matching method that prioritizes the selection of descriptors with a higher probability. The proposed matching method maintains the performance while also presents a speed improvement of approximately...
In this paper, we propose how to recognize upper-body poses using depth image based cylindrical coordinate system. In order to recognize the pose, we configure the cylindrical coordinate system belong to body features and the distance which is configured from camera to center of body using the pose candidate images. And we extract vectors of the features using depth information which are presented...
This paper presents the image analysis method for raw shiitake mushroom from image to shiitake grading to develop an automated system for grading a raw shiitake. Analysis pieces are the size of mushroom and the level of opening cap. Proposed method calculates the size of shiitake by using area sum of shiitake area and the rate of lamella area. To extract lamella area in image, proposed method uses...
Visual vocabulary is now widely used in many video analysis tasks, such as event detection, video retrieval and video classification. In most approaches the vocabularies are solely based on statistics of visual features and generated by clustering. Little attention has been paid to the interclass similarity among different events or actions. In this paper, we present a novel approach to mine the interclass...
This paper presents a novel scheme for human action recognition. First of all, we employ the curvature estimation to analyze human posture patterns and to yield the discriminative feature sequences. The feature sequences are further represented into sets of strings. Consequently, we can solve human action recognition problem by the string matching technique. In order to boost the performance of string...
We present a novel convex-optimization approach to solving the dense stereo matching problem in computer vision. Instead of directly solving for disparities of pixels, by establishing the connection between a permutation matrix and a disparity vector, we directly formulate the stereo matching problem as a continuous convex quadratic program in a simple, elegant and straightforward manner without performing...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.