基于复杂性指数的图像分割必要性判别技术  

Discriminatory technology of necessity for image segmentation based on complexity index

作　　者：魏庆[1] 卢照敢[1,2] 邵超[1] 

机构地区：[1]河南财经政法大学计算机与信息工程学院,郑州450002 [2]西安交通大学电子与信息工程学院,西安710049

出　　处：《计算机工程与应用》2013年第16期155-157,200,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61202285);河南省科技厅科学技术重点项目(No.12A510001)

摘　　要：考虑到图像分割的复杂性实质上与图像中可分割的区域个数相关,而图像分割的必要性是与图像中可分割的区域大小有关,针对图像分割实际应用中部分图像的内容较少、无明显语义,不必进行图像分割的情况,提出一种基于图像内容语义、图像分割复杂性的图像分割必要性判别测度。进一步基于其测度定义,进行了大量相关实验,实验结果表明,基于复杂性指数的图像分割测度很好地完成了预期的功能,能够成为图像分割必要性有效合理的衡量依据。Considering the complexity of image segmentation is essentially associated with the number of divisible area of image, and the necessity for image segmentation is related to the size of the dividable region in the image, in view of some of the image contents are less and lack of clear semantics, even unnecessary for segmentation in the practical application of image segmentation, an image segmentation necessity discriminatory measures based on image content semantics and the complexity of image seg- mentation is proposed. A lot of experiments based on the measure definition are carried out. Experimental results show that the proposed image segmentation based on complexity index measure well implements the expected function. The image segmenta- tion based on complexity index measure can be the effective and reasonable measure for the image segmentation necessity.

关 键 词：图像分割 复杂性 测度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]