基于一种改进的Hough变换的输电线提取与识别

在野外拍摄得到的输电线图像中，输电导线具有明显的线性特征，因此可以直线为基础来研究导线提取与识别的算法。Hough变换是经典的变换域直线提取算法，针对输电线图像的特点以及从Hough变换改进的思路出发，运用形态学分析和搜索聚类的方法先去除输电导线二值边缘图像中的背景噪声，提取出导线上的像素点。经过预处理后，再运用阈值区间的自适应估算方法来计算Hough变换的参数阈值，从而识别出图像中的输电线。实验结果表明，该方法具有较好的提取效果，并且能准确的识别出输电线，计算简单速度较快。

Extraction and Recognition of Transmission Line Based on an Improved Hough Transform

The transmission line image taken from field environment presents linear character obviously, so it can be regarded as a line to extract and recognize. The Hough transform is a general technique for identifying the location and orientation of certain types of feature in a digital image. From the properties of power lines and Hough transform, the morphological analysis and searching clustering method were used to remove the noise of power line edge image. And then, the power line pixel points were used to estimate the adaptive threshold interval according to certain rules. Finally, the power line was detected from these pixels using the obtained threshold interval. The experimental results show that the proposed method can extract and recognize the transmission lines with simple calculation and high speed.