Mean-gradient based two-dimensional histogram image segmentation algorithm in polar coordinate system

Traditional two-dimensional histogram image segmentation algorithms are susceptible to noise interference, poor segmentation accuracy and huge computational effort to find the best threshold. To solve these problems, this paper proposes a two-dimensional histogram image segmentation algorithm based on mean-gradient in-formation in polar coordinate system. Based on the traditional two-dimensional histogram, the image gradient information is introduced, and the mean and gradient values of the neighbouring pixels of the image are selected to form a two-dimensional histogram to increase the noise immunity and improve the segmentation accuracy. In addition, this paper maps the pixel points from the right-angle coordinate system to the polar coordinate system, and through the processing of noise and boundary points, the image pixel points are concentrated around 90°, so that the threshold segmentation can be performed only by the polar diameter information in the polar coordinate system, which greatly shortens the running time. Experimental results show that the algorithm improves segmentation accuracy and running speed in the case of more severe noise.