A Shadow Removal Algorithm with Gaussian Mixture Model
Tong Zhang · Jisuanji fangzhen · 2010
Shadow removal is an important part of Surveillance System,and it has great influence on object recognition and object tracking.Based on the improvement of the Gaussian Mixture Model,objects are detected.Gaussian Mixture Shadow Model is used for detecting the shadow.And the results are processed by Gaussian smooth filtering.The next main problem solved in this paper is to carry out binarization processing,select threshold value,and treat the result by morphology filter furthmore.Through the entire process concerned,the whole contour is obtained,shadow removal is processed easily,and the accuracy and reliability of object recognition and tracking is improved greatly.The experiment shows that this algorithm can detect moving shadows better,which has a good real-time performance and robustness.