摘要
Normalized Cut准则图像分割是基于图论的方法,是一种比较典型的规范化分割,但由于其运算量较大、收敛条件难以控制等缺陷,使得图像二值化分割不均匀。本文从Normalized Cut缺陷研究入手,利用遗传算法全局快速检索特性,改进Normalized Cut优化函数,获得精度理想的图像分割效果。
Normalized cut is a typical method of normalized segmentation based on graph theory. Some defects,such as large amount of computation and the uncontrollability of the convergence condition,make of the convergence binarization split.In this paper, we start with the research of Cut Normalized defect, and use genetic algorithm to improve the Cut Normalized optimization function, and improve the accuracy of the image segmentation results.
出处
《齐齐哈尔大学学报（自然科学版）》
2016年第3期25-28,共4页
Journal of Qiqihar University(Natural Science Edition)
基金
2016年度安徽省教育厅人文社科重点研究项目"安徽国家级数字出版基地政策工具选择研究及效果评估"
