基于遗传算法的图像阈值分割的研究
2016-04-11周爱霞李春贵陶佳伟
周爱霞+李春贵+陶佳伟



摘要:图像阈值分割技术在图像分析和图像识别中有着重要的意义。目前在实际应用领域中,现有的图像阈值分割算法存在着如目标轮廓模糊,耗时性,或一些重要细节被忽视掉等问题。针对这些问题该文展开了研究,重点研究了图像阈值分割法中经典算法最大类间方差(Otsu)算法。为了解决 Otsu 算法用时长这一问题,该文结合遗传算法以及它的特点,将遗传算法用于Otsu的图像分割方法中并对阈值进行寻优,加快了算法的收敛速度。实验结果表明,该文的算法具有良好实时性、分割效果好的特点。
关键词:遗传算法;Otsu算法;图像分割;阈值
中图分类号:TP18 文献标识码:A 文章编号:1009-3044(2016)04-0198-04
Study on Image Threshold Segmentation based on Genetic Algorithm
ZHOU Ai-xia1, LI Chun-gui 1,2, TAO Jia-wei 1
(1. Guangxi University of Science and Technology, College of Electrical and Information Engineering, Liuzhou 545006, China;2.Guangxi University of Science and Technology, Computer Science and Communication Engineering Institute, Liuzhou 545006, China)
Abstract: Image thresholding segmentation is of importan significance for image analysis and recognition. Presently, there are several problems in the existed thresholding segmentation algorithms when applied in the practical application, such as time consuming, the vaguely, or important details are ignored. Aimed at these problems, this dissertation expands a series of researches, the typical threshold segmentation algorithm, Otsu, is chosen as the center point of the research. In order to overcome time-consuming of the Otsu algorithm better, Combining genetic algorithm with its characteristics, let it applied to Otsu image segmentationAlgorithm to find the best threshold, improved the convergence rate of the algorithm.The experimental results show that the new algorithm is not only higher quality but also possess a well real-time.
Key words: the genetic algorithm; Otsu method; image segmentation; threshold
1 概述
图像分割其应用在很多领域都十分广泛,并且涉及了各种不同类型图像,它在图像处理、人工智能等多个领域中非常重要,是计算机视觉中关键的步骤[1]。对计算机视觉中的图像理解的关键要看图像分割结果的好坏。图像分割技术主要有区域分割技术、阈值分割技术、直方图法和边缘分割技术等。
图像阈值分割的关键是选取一个适当的阈值,把图像分成灰度级不同的两部分(目标区和背景区),从而判断出图像中的每个像素点是在属于哪个部分,进而产生了对应的二值图像[2]。目前,现有的阈值分割算法,如直方图阈值分割、迭代式阈值分割、Otsu阈值分割等[3]。Otsu法是基于整幅图像的统计特性,来实现图像阈值的选取,分割效果良好,但它对目标大小和噪声十分敏感,其运算量大,运行速度缓慢影响了其应用的实时性。……
