For the project intelligent video surveillance systems, the primary objectives are video understanding such as target detection, target tracking, human behaviours activity understanding, and video representation such as video compression, video resolution enhancement. The project aims to achieve these tasks of understanding and representation with respect to surveillance video. The algorithms and methodologies developed are steps towards an automatic or semi-automatic surveillance system. The project proposed several schemes in video understanding such as human activity segmentation, human detection using shape prior, and in video representation such as object based video coding, video super-resolution for face enhancement.