It is well known that solar tracking systems can increase the efficiency of exiting solar panels significantly. In this paper, a hybrid solar tracking system has been developed by using both astronomical estimates from a GPS and the image processing results of a camera vision system. A decision making process is also proposed to distinguish current weather conditions using camera images. Based on the decision making results, the proposed hybrid tracking system switches two tracking control methods. The one control method is based on astronomical estimates of the current solar position. And the other control method is based on the solar image processing result. The developed hybrid solar tracking system is implemented on an experimental platform and the performance of the developed control methods are verified.