融合灰度和非局部空间灰度特征的二维Otsu法  

Two-dimensional Otsu's method based on gray and non local spatial gray feature

作　　者：赵凤[1] 范九伦[1] 

机构地区：[1]西安邮电学院通信与信息工程学院,陕西西安710121

出　　处：《西安邮电学院学报》2012年第3期10-14,共5页Journal of Xi'an Institute of Posts and Telecommunications

基　　金：国家自然科学基金资助项目(61102095);陕西省教育厅科研计划资助项目(11JK1008)

摘　　要：为了克服图像噪声对二维Otsu阈值分割方法性能的影响,采用图像像素的灰度和非局部空间灰度特征构造新的二维直方图,其中像素的非局部空间灰度特征是通过对与当前像素具有相似邻域结构的像素灰度加权平均得到的。将此直方图引入到快速二维Otsu阈值分割方法中,可得融合灰度和非局部空间灰度特征的二维Otsu阈值分割法.实验结果表明改进算法对图像噪声具有一定的鲁棒性,在含噪图像上的分割结果比较理想。Aiming at overcoming the influence of the image noise to the performance of two-dimensional Otsu＇s thresholding image segmentation, a novel two-dimensional histogram based on the gray and non-local spatial gray feature is constructed in this paper. The non local spatial feature of the pixel is extracted from a set of pixels with a similar neighborhood configuration of the current pixel. Then this novel histogram is introduced into the fast two-dimensional Otsu＇ s thresholding segmentation algorithm and the two-dimensional Otsu＇s thresholding segmentation method based on gray and non local spatial gray feature is proposed. Experimental results show that our method is robust to image noise and obtains satisfying segmentation results on the noisy images.

关 键 词：非局部空间灰度 二维直方图 OTSU 阈值化 图像分割 

分 类 号：TN911[电子电信—通信与信息系统]