基于二维Otsu阈值修正的图像分割新方法

首先采用对二维直方图斜分割和查表的方法，解决了传统二维Otsu方法分割图像计算耗时，难以实时实现的缺点。其次基于灰度统计的思想，针对二维Otsu法处理小目标图像难以实现正确分割的缺点，提出了一种在实现过程中采用迭代的阈值修正新方法。最后设计了一种新型滤波器对分割后的图像进行滤波降噪处理。实验结果表明，阈值修正后的二维Otsu改进算法对小目标图像分割效果明显，而且新型滤波器对滤除散布在目标与背景中的噪声非常有效。将阈值修正法和新型滤波器结合使用，不仅快速，而且准确，取得了良好的分割效果。

New Image Segmentation Method Based on Threshold Correction of Two Dimensional Otsu

First with the shortcomings of time consuming and being hard to real time realize,oblique segmentation of 2D histogram and lookup table were adopted to resolve traditional 2D Otsu method’s disadvantages.Then a new method based on probability was proposed to resolve that 2D Otsu was not applied to small target images.The method corrected the threshold and iterative method was used in realization.A new filter was designed to filter the image segmented in the end.The results show that the threshold correction is applied to images including small targets,and the result of segmented images is very good.At the same time,the new filter is effective on noise spread in targets and backgrounds.When threshold correction and new filter are used together,it is not only fast,but also correct,and the segmented result is favorable.