基于最优阈值的叶类中药边缘检测算法研究  

A study on edge detection algorithm of leaf traditional Chinese medicine based on optimal threshold

作　　者：侯青[1] 颜鹏[2] 孙静[1] 任娜娜[2] 刘玉娥[2] 

机构地区：[1]陕西中医学院,陕西咸阳712046 [2]长安大学,西安710064

出　　处：《信息技术》2014年第2期69-72,共4页Information Technology

基　　金：陕西省教育厅专项科研计划项目(12JK1029);陕西省科学技术厅专项科研计划项目(2013SF02-7)

摘　　要：提出了一种利用最优阈值分割和形态学处理相结合的边缘检测方法,首先通过Otsu最优阈值算法分割数字图像,然后利用形态学腐蚀运算腐蚀图像,最后利用分割后图像与腐蚀过的图像的相减获得边缘。该方法实现了对叶类中药显微图像效果更优的边缘检测,同时对比了传统边缘检测算法对叶类中药显微图像的图像边缘检测效果,为下一步的叶类中药显微图像的分割与自动识别奠定基础。This paper presents an optimal threshold segmentation and morphological processing combined with edge detection method, the optimal Otsu threshold segmentation algorithm of digital image, and then uses morphology erosion corrosion image, finally gets the edge using subtraction segmentation image and the image of the corrosion. The method realized the traditional Chinese medicine image better edge detection, and compared to the traditional image edge edge detection algorithm on Chinese medicine image detection results, to lay the foundation for the next step of te microscopic image of leafy Chinese medicine segmentation and recognition.

关 键 词：最优阈值分割 形态学处理 叶类中药 显微图像 边缘检测 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]