一种改进的细胞图像分割迭代算法  

An improved iterative algorithm for cell image segmentation

作　　者：李骜 李一兵[1] 孟霆 侯欣彤[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院

出　　处：《黑龙江大学自然科学学报》2010年第6期811-813,817,共4页Journal of Natural Science of Heilongjiang University

基　　金：国家自然科学基金资助项目(60703090);哈尔滨市科技创新人才研究专项基金项目(2008RFQXG025)

摘　　要：图像分割是图像轮廓提取,纹理特征提取和形状特征提取等图像分析的重要基础。它也直接影响着图像更高级别的处理质量。提出一种改进的图像分割迭代算法,通过在一般的迭代算法中添加第二门限,以达到更精确的将各像素划分到所属区域降低误判概率的目的。在分析的基础上,给出了该算法与一般迭代分割算法的对比图像。为了观察分割效果,还给出了应用分割图提取边缘与Canny算子的对比图像。其结果均表明,噪声干扰下,改进算法的分割过程更为准确。Image segmentation is an important basis for Image analysis such as image contour extraction,texture and shape feature extraction,with a direct impact on the image quality for a higher level of treatment.The improved iterative algorithm is provided for image segmentation,adding the second threshold in the general iterative algorithm,in order to achieve a more accurate division to allocate each pixel into their respective regions,for the purpose of reducing the probability of miscarriage of justice.On the basis of the analysis,the contrast images is given between the improved algorithm and the general iterative algorithm.For observing the effect expediently,it also gives the contrast images between the contour extraction by segmentation image and the canny operator.The results showed that when the image subject to the noise interference,the segmentation process of improved algorithm is more veracious.

关 键 词：图像分割 迭代算法 第二门限 

分 类 号：TN929.5[电子电信—通信与信息系统]