Video Content Representation as Salient Regions of Activity
نویسندگان
چکیده
In this paper, we present a generic and robust representation of video shots content expressed in terms of salient regions of activity. The proposed approach is based on salient points of the image space, thus minimizing the computational effort. Salient points are extracted from each frame. Their trajectories are computed between successive frames and the global motion model is estimated. Moving salient points are selected from which salient regions are estimated using an adaptive Mean-Shift process, based on the statistical properties of the point neighborhoods. The salient regions are then matched along the stream, using the salient points trajectories. The information carried by the proposed salient regions of activity is evaluated and we show that such a representation of the content forms suitable input for video content interpretation algorithms.
منابع مشابه
A Bottom-Up Spatiotemporal Visual Attention Model for Video Analysis
A video analysis framework based on spatiotemporal saliency calculation is presented. We propose a novel scheme for generating saliency in video sequences by taking into account both the spatial extent and dynamic evolution of regions. Towards this goal we extend a common image-oriented computational model of saliency-based visual attention to handle spatiotemporal analysis of video in a volume...
متن کاملFinding spatio-temporal salient paths for video objects discovery
Many videos capture and follow salient objects in a scene. Detecting such salient objects is thus of great interests to video analytics and search. However, the discovery of salient objects in an unsupervised way is a challenging problem as there is no prior knowledge of the salient objects provided. Different from existing salient object detection methods, we propose to detect and track salien...
متن کاملA Novel Approach to Background Subtraction Using Visual Saliency Map
Generally human vision system searches for salient regions and movements in video scenes to lessen the search space and effort. Using visual saliency map for modelling gives important information for understanding in many applications. In this paper we present a simple method with low computation load using visual saliency map for background subtraction in video stream. The proposed technique i...
متن کاملVideo Representation via a Continuous Probabilistic Framework
In this work we describe a novel statistical video representation and modeling scheme. Video representation schemes are needed to enable segmenting a video stream into meaningful video-objects, useful for later indexing and retrieval applications. In the proposed methodology, unsupervised clustering via Guassian mixture modeling extracts coherent space-time regions in feature space, and corresp...
متن کاملHuman-Content and Gesture-Event Video Coding
Currently, bandwidth limitations pose a major challenge for delivering high-quality multimedia information to users. In this research, we aim to provide a better compression of human-centered video sequences such as lectures, monologues, and presentations. Based on the idea that people pay more attention to face and hand regions in videos containing people speaking, our approach encodes those r...
متن کامل