基于局部增强和阈值分割的芯线彩色图像分割算法

摘要

不同颜色芯线的有效识别是制约数据线生产自动化的关键因素之一;阈值分割是一种简单有效地图像分割算法, 但对于颜色信息较多的情况鲁棒性较差;基于指数函数的增强图像可以有效地放大和突出感兴趣的图像区域;本文提出了一种结合局部增强和阈值分割的分割算法, 通过对图像进行局部增强以放大感兴趣灰度区域, 然后再通过阈值分割的方式筛选出目标区域;该方法能够快速有效地分割出多种颜色区域, 也具有良好的鲁棒性.%The identification of different colors of core lines is one of the primary restrictions of automation in dataline production.The simple and effective thresholding technic is not robust enough in detecting various colors with tiny differences.An image enhance process based on exponential function can effectively highlight the region of interest.A method that coupled image enhancement with thresholding was proposed in this paper.The object region was extracted by thresholding when the image had been locally enhanced.The experiments showed that the proposed method was effective and robust enough in distinguishing the various colors of core lines.