The purpose of this dissertation was to design a multiple object tracking system which was utilized to evaluate sperm quality of pig insemination. In the study, we applied the multiple Kalman filter and tracking correspondence algorithm to solve rotation , disappear and occlusion problem of sperm, and estimated activities of sperm. In sequential images as the results shown, the developed system could not only calculate the locus and activities of pig sperm, but also provide a more efficient and precise method than artificial estimation.