机器视觉中图像分割问题研究  

Study on Image Segmentation in Machine Vision

作　　者：侯远韶 张瑶 

机构地区：[1]郑州工业应用技术学院信息工程学院,河南郑州451100

出　　处：《洛阳理工学院学报（自然科学版）》2016年第2期76-80,共5页Journal of Luoyang Institute of Science and Technology：Natural Science Edition

基　　金：河南省科技攻关计划项目(0721002210032)

摘　　要：随着机器视觉技术的普遍应用,图像的采集、处理越来越受到人们的重视。图像分割作为图像处理的重要一环,影响着图像预处理的效果,同时也是后续图像进行分析、理解的基础。现有的图像分割技术没有一个统一的理论体系,不能够根据不同的图像特征选择合适的分割算法。本文分析了现有图像分割技术的原理及应用,在阈值分割技术的基础上,提出了基于模糊阈值的自适应背景区域分割方法。实验表明这种方法在不影响图像理解精确度的前提下,能够有效提高算法的使用范围。With the universal application of machine vision technology,image acquisition and processing have been paid more and more attention. Being the important part of image processing and the basis of understanding and analysis of the subsequent image,image segmentation has effect on the results of image prepossessing. But there is not a unified theoretical system for the existing image segmentation technology to guide us choosing a suitable algorithm according to the different image characteristics. By analyzing the principle and application of the existing image segmentation technology,an adaptive background region segmentation method based on fuzzy threshold is presented in this paper. The experimental results show that the method can effectively extend the application scope of the algorithm without affecting the accuracy of image understanding.

关 键 词：机器视觉 图像分割 自适应技术 模糊阈值 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]