摘要
对于图像测量中平行线边缘点的拟合问题,本文通过聚类分析法中的DBSCN算法去除边缘点图像中的噪声数据,并结合最小二乘法对平行线边缘点进行分类拟合,解决了传统算法中根据图像特点人为划分区域进行分别拟合的问题,达到了机器对图像进行自动处理的目的.
For the parallel edge points fitting in image measurement, this paper combines with the least square method to classify parallel edge points fitting through clustering analysis algorithm to eliminate the noise in the edge points image. The traditional algorithm based on image characteristics dividing area artificially to fitting respectively is solved. The result achieves the purpose of automatic image processing by machine.
出处
《计算机系统应用》
2014年第6期175-178,共4页
Computer Systems & Applications
关键词
图像测量
最小二乘法
聚类分析
image measurement
least squares
clustering analysis
