In addition to the above methods, the researchers also proposed other methods such as Markov random field, direction projection, maximum stable extremum region detection and so on. These methods have their own advantages and disadvantages in terms of complexity and operating efficiency. For example, the directional projection method achieves rough positioning of human targets and separation of adhesion regions by summing the gray values of foreground pixels along the main axis and thresholding them, which has the advantages of simplicity and efficiency, but it cannot handle human targets with complex shapes . The maximum stable extremum area detection method can detect the bright flat area in the image, but to obtain the complete human body area, it needs to use the prior knowledge of human body shape to do a lot of filtering and merging of the initial detection results.