<![CDATA[For surveillance system, the background subtraction plays an important role for moving object detection with an algorithm embedded in the camera. Since the existence algorithms cannot satisfy the good accuracy on complex backgrounds including illumination change and dynamic objects, we have put forward the concept of Common Vector Approach (CVA) as a new idea for background modelling. Effectiveness of proposed method is presented through the experiments on popular Wallflower dataset. The obtained visual outputs are compared with well-known methods based on the subjective and objective criteria. From the overall evaluation, we can note the proposed method is not only exhibit successful foreground detection results, but also promises an effective and efficient system for background modelling.]]>
Copyrights © 2016
