摘要
设计了一个扩展Robert算子,该算子能够在有噪声的情形下对图象中各种宽度的线条进行检测。当使用Hough变换确定直线的参数时,首先只用Hough变换检测第一条最为显著的直线,随后去掉该条直线及其附近的点,然后再次对图象进行Hough变换,并重复此过程,直到找到所有直线或者Hough变换后参数平面上的值都小于某个阈值为止。使用此改进后的Hough变换能够准确地检测到图象中构成网格的直线的参数。给出了具体的检测例子。
An extended Robert arithmetic operator is designed,which can detect lines with different width in images with noise.When determining the parameters of the lines,at first only obtain the most obvious line in the image with ordinary Hough Transform,then delete the line and the pixels near the line in the image and detect the second line in the image with Hough Transform.Repeat this processing until all lines in the image are detected or the value of Hough Transform is very small.With this improved Hough Transform,the parameters of the lines constructing the grid in the image can be obtained accurately.An application example is given.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第20期64-66,共3页
Computer Engineering and Applications
基金
国家自然科学基金重点项目(编号:60234030)