Removing Blindness Shadows from Image Based on Its Points Calculation in High Dimensional Space

Shadow elimination is critical for robust and reliable pattern recognition in vision system. Because cast shadows have most motion features as the moving objects, it is difficult to eliminate it. This paper presents a novel moving cast shadow elimination method. Different from traditional methods which only consider algebra translation in shadow modeling, this method introduces a method from the point of view points in high dimensional space. Firstly, we use special kernels to generate some image points from original image; then the most similar one which has little shadow affection is found; lastly, an adjustment method is proposed to further eliminate shadow in image from the point of view high dimensional space. The experiments demonstrate shadow region can be eliminated quite well.