一种基于改进SUSAN算子的钨矿初选算法  

A Primary Algorithm of Tungsten Ore Based on SUSAN Operator

作　　者：肖玲玲[1] 谢斌[1] 赵秀鸟[1] 

机构地区：[1]江西理工大学信息工程学院,江西赣州341000

出　　处：《江西理工大学学报》2011年第3期30-32,共3页Journal of Jiangxi University of Science and Technology

基　　金：江西省教育厅青年科学基金资助项目(GJJ11132)

摘　　要：以改进SUSAN算子为核心的钨矿石初选算法是一种针对性较强的算法.此算法采用改进的SUSAN算子的边缘提取方法,利用形态学开运算估计背景,与原始图像全局灰度进行几何运算来减弱背景的复杂纹理和噪声对图像的影响,以此来改善目标边缘提取的效果,并结合SUSAN基于图像灰度相似性比较的边缘检测算法.它不需梯度的计算,具有算法简单、定位准确、抗噪声能力强等特点,提高了算法效率,实现二者的优势互补,提高图像边缘检测的效果.Tungsten primary machine to achieve automatic sorting system is the key technology of tungsten ore image featureextraction,identification and sorting algorithm.SUSAN operator to improved fast algorithm as the core of tungsten ores and primary algorithm is a targeted optimum algorithm.A primary algorithm of tungsten ore based on SUSAN operator is designed and realized.The algorithm adopts the improved SUSAN edge extraction operator with morphological opening operation to estimate background and global gray image taking the original geometry of complex operations to weaken the background noise on the image texture and the influence.In order to improve the effect of edge detection,the algorithm is combined with the similarity SUSAN comparison based on image gray edge detection algorithm.The algorithm,which does not require gradient calculations,is simple,accurate in positioning and antinoise.The effect of edge detection is enhanced.

关 键 词：钨矿初选 特征提取 SUSAN算子 

分 类 号：TP399[自动化与计算机技术—计算机应用技术] TD952[自动化与计算机技术—计算机科学与技术]