Shadow Detection Method Based on Gradient Feature
Xinzhong Zhu · Computer Technology and Development · 2009
In the traffic video vehicle detection,accurately obtaining vehicle is one of the most important problems.Since the shadow exists in most cases,and affects traffic information inspection accuracy,therefore,judging the shadow is very important.Use Surendra algorithm to extract background,then use background difference method to extract moving foreground,and then enhance foreground area.At last,using gradient operator calculate gradient.According to the gradient of moving vehicles is different to gradient of relevant background,while gradient of moving shadows is similar to that of relevant background,to detect and remove the shadow.The results of experiment prove that this method can effectively remove shadow,improve the accuracy of the vehicle detection.This method is simple,and effectively reduces the computational cost,and robust to illumination changes.