Abstract
In this study, we have developed a new automatic animal motion activity measurement system based on the computer image-processing techniques, and this system could be applied to animal locomotion recorder, asymmetry rotation and travel track analysis. This system uses webcam as the real time image source or pre-saving video files for analysis. Real time video stream input from webcam were compared and analyzed through each successive image frame to separate the object from the background, and computed the motion parameter from the changes and variations of object images. The contours of objects are extracted by first searching the objects through Background Subtraction, and then Region Growing method and fast contour chain algorithm are employed to reduce the computational complexity for real-time application. The principle axis of object is used to measure the rotation of rats, and the moving tracks of rats are recorded into a file system. The simple and economical hardware requirements of this automated observation system result in the advantages of lower cost, but it can provide convenient setup with a high efficiency and high accuracy data correction. We expect this new system to be profitably employed for those studies on normative, pharmacology or neuropathology behavioral activity studies.
Keywords: Parkinson’s disease, automatic motion measurement, object tracking, object segmentation, contour extraction