The Probabilistic Hough Transform (PHT) is the most active research subject in the Hough field in recent years. However there are quite a number of questions to be solved out. e.g., How to determine the stopping rules? For a given pattern to be detected, which parameterization should be used? How to eleminer (reduce) the noise in Hough space coming from image noise? How to effectively deal with the objects' interferences? Based on which strategy Hough space should be discretized? The project is intended to find out answers to these questions.