Video object extration is one of the most important areas of video processing in which objects from video sequences are extracted and used for many applications such as surveillance systems, pattern recognition etc.
In this research work, an object-based technique based on the spatiotemporal independent component analysis (stICA) is developed to extract moving objects from video sequences. Using the stICA, the preliminary source images containing moving objects in the video sequence are extracted. These images are processed using wavelet analysis, edge detection, region growing and multiscale segmentation techniques to improve the accuracy of the extracted objects. A novel compensation method is applied to deal with the nonlinear problem caused by the application of the stICA directly to the video sequences. The recovered objects are indexed by the singular calue decompensation (SVD) and linear combination analysis. Simulation results demonstrate the effectiveness of the stICA-based object extraction technique in content-based video processing applications.
A novel approach for object extraction from video sequences based on continuous background/foreground classification
