Abstract:The traditional two-dimensional image thresholding segmentation algorithms exist some shortcomings that are the area division of two-dimensional histogram (Part of the target points and background points is divided into edge points or noise points, while part of the edge points and noise points is divided into the target point and background points) and high time complexity of searching the best threshold vector, a new two-dimensional histogram is proposed by using human vision model, and a new region division method about two-dimensional histogram is proposed, and the same time image thresholding segmentation based on human vision model and maximum entropy is proposed, the proposed image thresholding segmentation algorithm reduces the time complexity and has good segmentation performance. According to some evaluation standards for image segmentation result, a series of experiments show the proposed algorithm has better segmentation effect compared with several typical two-dimensional threshold segmentation algorithms.