Motion sensed video storage algorithm for surveillance recording

In this paper, implementation of a real time motion detection algorithm along with the video compression technique using MATLAB Simulink Software described. Motion Detection algorithm is used to start of video recording only when motion is detected in the frames which reduces temporal redundancy present in the sequence of input frames. Detection of changes in a background frame initiates video recording with real time video compression algorithm to reduce the size of memory. Motion Detection algorithm is implemented using Mean Reduced Sum of Average Difference Method (MRSAD) and for Video compression technique Two Dimensional Discrete Cosine Transform is used which takes advantage of statistical behaviour of video data in spatial domain. In this way high degree of compression is achieved in both spatial and temporal domain without using motion estimation and compensation techniques.