Shadow Elimination Method for Video Surveillance


  •  Huaiqiang Liu    
  •  Feng Guo    

Abstract

With regard to the weakness and shortage of traditional moving object segmentation method, this paper presents an effective segmentation method for moving objects in video surveillance. The difference image of color distance which is between current image and the reference background image in RGB color space is first obtained. According to the mono-modal feature of histogram of the difference image, an adaptive clustering segmentation method based on histogram is proposed. The morphology filtering is employed to remove the noise existing in the segmented binary image. An updating scheme for background image is introduced to follow the variation of illumination conditions and changes in environmental conditions. In order to remove unwanted shadows of moving regions, an efficient multi-object shadows distinguishing and eliminating method for surveillance scene was presented in this paper. Experimental results show that the proposed method is simple and effective for moving object segmentation and eliminating shadows.



This work is licensed under a Creative Commons Attribution 4.0 License.