基于改进遗传算法的阈值图像分割方法  

Research on Threshold Image Segmentation Method based on Improved Genetic Algorithm

作　　者：李茂民 邹臣嵩[1] LI Maomin;ZOU Chensong(Department of Electrical Engineering,Guangdong Songshan Polytechnic,Shaoguan 512126,China)

机构地区：[1]广东松山职业技术学院电气工程学院,广东韶关512126

出　　处：《软件工程》2022年第1期37-40,共4页Software Engineering

基　　金：韶关市科技计划项目(2019sn094,200811224533986).

摘　　要：本文提出一种新的基于改进遗传算法和阈值图像分割相结合的人像图像分割方法。这种新的改进方法以遗传算法为基础,利用遗传算法具有较高的搜索效率、明显的搜索精度,提升了图像分割阈值的精度获取,提高了图像分割的抗噪能力,在提升阈值稳定的同时,提升了阈值的获取速度及获取精度,解决了传统算法应用于人像图像分割时分割效果不理想、分割精度较低的缺点。经过实验验证,利用本文改进算法能达到较好分割效果,具有较好的抗噪能力,从而缩短分割图像时间。This paper proposes a new portrait image segmentation method based on a combination of improved genetic algorithm and threshold image segmentation.Genetic algorithm,which has higher search efficiency and obvious search accuracy,is used as a basis in the new improved method to improve the accuracy acquisition and the anti-noise ability of image segmentation.While increasing the stability of the threshold value,it improves the speed and accuracy of the threshold value acquisition,so to overcome the shortcomings of unsatisfactory segmentation effect and low segmentation accuracy when traditional algorithms are applied to portrait image segmentation.Experiments verify that the improved algorithm in this paper achieves a better segmentation effect and has better anti-noise ability,thereby shortening the image segmentation time.

关 键 词：人像 图像分割 遗传算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]