Robust adapted object detection under complex environment

In this paper, we present a novel robust technique for background subtraction in different complex conditions (e.g. sudden illumination changes, swaying leaves, and camera vibrations). Unlike the previous works, the proposed method utilizes multiple point pairs that exhibit a stable statistical intensity relationship as a background model. The intensity difference between pixels of the pair is much more stable than the intensity of a single pixel, especially in varying environments. Furthermore, our proposed method focuses more on the history of global spatial correlations between pixels than on the history of any given pixel or local spatial correlations. we also adopt an adapted judgement criterion to ensure our method displays well in real-time detection. The approach has been compared with the state of the art on videos from several challenging datasets (PETS, Wallflower, and i-Lids), demonstrating that superior object detection is achieved.