Mining Moving Object Attributes Based on Pixel Location for Fixed Camera

This paper presents the concept of extracting information from a moving object using a fixed camera in an enclosed environment in order to extract some of the image attributes. The work relies on tracking a single object in an indoor environments and the knowledge extraction of a particular object is based on three preset measurement tools. The first tool is pixel location retrieval by calculating the center of mass for the object of interest. Second attribute extraction is based on the time spend by object of interest in each frame and in each predefined zones. Finally the third measurement approach is achieved by calculating the pixel frequency distribution of the object. This work form a precursor to identifying intruder (or any interested moving object) based on pixel activity using fixed angle view video images.