The embodiment of the invention discloses an
image processing method, device and
server, which are used to improve the
coincidence degree of a target frame and a foreground target. In the embodiment of the present invention, first obtain the input frame and the background frame, then determine the row or column to be
cut in the target frame, calculate the gradient difference matrix of the input frame and the background frame in the row or column to be
cut, and calculate the gradient difference matrix The average value of the gradient difference, judge whether the average value is less than the gradient comparison value, if it is less than,
cut off the position to be cut from the target frame. In the embodiment of the present invention, the similarity between the input frame and the background frame is used as the basis for
cutting the target frame, and the average value of the gradient difference value of the row or column in the target frame of the input frame and the background frame is respectively calculated as a comparison feature, and compared with the gradient comparison value , so as to judge whether the current row or column is a real foreground target, if not,
cut off the current row or column from the target frame. In this way, the target frame can be effectively corrected, and the
coincidence between the target frame and the foreground target can be improved.