基于形态学与二维Otsu相结合的红外图像分割方法

摘要

Otsu法是常用的基于阈值的图像分割方法之一,二维Otsu法利用图像的像素灰度值分布及邻域像素的平均灰度分布构成的二维直方图对图像进行阈值分割.由于红外图像具有低对比度、低信噪比、边缘模糊的特点,仅采用二维Otsu对其进行分割,由于噪声的影响,分割后的图像会有边缘信息不清晰以及误分割的问题.针对这个问题,本文提出的形态学与二维Otsu相结合的红外图像分割方法.实验证明,利用形态学可以保留图像基本形状,弥补分割图像细节,并使图像的轮廓更光滑的特点,达到较好的红外图像分割效果.%Otsu's method is one of the frequently used thresholding segmentation methods,the two-dimension Otsu takes twodimensional histogramimage consists of image pixel gray value distribution and the average gray level distribution of neighborhood pixels into consideration.Because of the infrared image has some characteristics such as low contrast、low signal-to-noise ratio 、edge is fuzzy.If only use two--dimension Otsu for image segmentation,the edge information of the image been segmented is not clear as well as the error segmentation because of the noise.To overcome this problem,this paper presents an improved infrared image segmentation based on the combination of morphological and two-dimensional Otsu method.Experiments show that by using morphology can keep image basic shapes,make up for the segmentation of image details,and make the image contour more smooth,the Infrared image segmentation obtain better results.