The system use surveillance camera and image rendering developed a new solar positioning system. The system’s position detection method is to use surveillance cameras captured source image, and then use rendering method to determine the location of the light source to adjust the orientation of solar cells to improve the utilization of the battery. We use existing camera filming the scene of the source image, and then the light source position will be determined. Through the back of our tracking software processing, adjusting the orientation of solar collectors, will allow maximum utilization of solar energy and provide solar tracking areas a new and reliable detection methods. This article focuses on algorithm design of the source tracking software. This article uses the prior knowledge of three-dimensional objects in the photo image to model. Priori knowledge can help us to determine all the parameter except light source location, we full traversal all the possible light source position space. After getting the rendering image, rendered image and source image will be compared through the evaluation function, thus we will get the best source location parameters.