A Spatio-Temporal Framework for Moving Object Detection in Outdoor Scene

This paper addresses the problem of video object detection under illumination variation condition. Since it is a very general case in outdoor environment, hence many attempts have been made to design a robust and efficient algorithm, which takes care of any such case of illumination variation. In this paper we have proposed an effective spatio-temporal framework based algorithm which computes the inter-plane correlation between three consecutive Red, Blue and Green planes of three consecutive video sequences by using a correlation function. The correlation matrix obtained is then used to construct an image which gives a rough estimate of the object to be detected. This image is then fused with the moving edge image in a deterministic framework to detect the final moving object in the video. The algorithm is tested in different outdoor and indoor situations and found to be very much efficient in terms of the misclassification error.