A Correlation tracking algorithm based on the function of the minimum average absolute difference is studied. It aims to search the biggest corresponding parts of the minimum average absolute difference in the next frame image using pixel gray values of part of image region as a template to match and solve the target motion parameters. The matching criterion is small computation, simple operation and can reduce the computing time, improve the algorithm running efficiency. The test to track the car running on the road using this tracking algorithm is presented. The experimental result shows that this algorithm can be very good to track the moving car and the magnitude of tracking window deviating from the moving car is very small.