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.
Considering the requirement for a more objective result of image retrieval, the features of image itself attract more and more attentions of researchers. Recently, the shape context of images is widely applied in image matching for its good invariance of translation and scale, as well as its good robust for small geometric distortion. Thus, the features of shape context are introduced in this paper...
We present a multi-feature based object extraction algorithm in this paper. Cutting in from the angle of multi-feature, this algorithm combines the high-efficient and good performance Graph Cut framework. We add foreground shape information and motion estimation as a compensation of the instability of using single color information.
Video caption extraction has become a very popular research area in the last few decades. Many reasons makes it a challenging task. A large number of techniques have been proposed to address this problem. This paper reviews the progress in this area and various methods towards different stages of text extraction in videos, and also discusses the promising direction of the future research.
GrabCut is a classical efficient algorithm used in image segmentation. In this paper we provide an approach of video segmentation—an algorithm which is called “GrabCut in local window algorithm” based on GrabCut algorithm. First, make cutout with GrabCut in series of overlap local windows and get corresponding local contour boundaries. Then the contour boundaries in local windows are linked to get...
This paper researches the features of pornographic videos sensitive body videos, and presents a method to recognize and shield the sensitive content automatically. Skin color is one of important cues for pornographic video's detection. Firstly, transform the color space, calculate Gaussian probability distribution, definite threshold value, analyze texture and noise to extract skin message from a...
A fused background model that combines the eigenbackground with Gaussian models is proposed. We adopt the eigenspace model to build the intensity information for each pixel. Unimodal Gaussian density methods with less computational cost are used to describe color information for each pixel. An adaptive strategy is used to integrate the two models. Using the fused background model, we subtract the...
We introduce the first visual dataset of fast foods with a total of 4,545 still images, 606 stereo pairs, 303 3600 videos for structure from motion, and 27 privacy-preserving videos of eating events of volunteers. This work was motivated by research on fast food recognition for dietary assessment. The data was collected by obtaining three instances of 101 foods from 11 popular fast food chains, and...
This paper proposes background segmentation methods that integrate color, Ibp contrast and motion cues so as to improve the segmentation results. The background color models based on GMMs are developed to describe the scene. Motion cues is presented to detect the legitimate moved foreground pixels combining with the background subtraction. Lbp contrast features are used to calculate the image contrast...
In this paper, we present a biased sampling strategy for object class modeling, which can effectively circumvent the scene matching problem commonly encountered in statistical image-based object categorization. The method optimally combines the bottom-up, biologically inspired saliency information with loose, top-down class prior information to form a probabilistic distribution for feature sampling...
In this paper, we propose a novel framework of object categorization, namely layered object categorization, which takes advantage of hierarchical category information and performs object categorization at different levels. The proposed hierarchical structure of object categories is built bottom-up and top-down simultaneously accordingly to cognitive rules. First, part-based models are learnt to evaluate...
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.