Research Of Power Lines Recognition In Transmission Line Image Based On An Improved Hough Transform

With the development of information and technology, all walks become more and more intelligent and modernization, also the power line inspection. In recent years, researchers from all over the world do a lot of works to explore this province combining the aerial surveying technique and digital image processing, whatever in practical or theory. Although our human beings works cannot be replaced now, it will come true in the future through ours efforts.In this paper, through analyzing and comparing the main content of related research papers, the power line recognition in transmission line image will be the key point of this researching. By calculating the position of the power line from the image, a method is introduced to confirm the camera can take pictures clearly. It gives some theoretical basis to design the image taking system.The quality of the transmission line image can be improved by image processing. As the gray transformation is used to simple the original RGB data, optical correction and wave filtering is used to clear the line images. After that, morphology analysis and edge detection as the method used to achieve image segmentation. Combining with the easy calculate method to delete the false parameter, and extract the transmission line from the image.According the characters of the power lines from the image, also comparing several normal technologies of straight line detection, the Hough transform will be chosen to detect the power lines from the transmission line image. An improved method is studied to get the high speed and calculate less in Hough transform by optimizing the procedure of searching the peak point.Finally the power line recognition technology can help camera system adjust the environment to take the picture we need, such as broken or rusting etc. An algorithm is presented to get the relationship between the line positon in the image and camera position stop or move. If the power lines keep in the center of this image, the video can record clearly. There are also some images which be used in the experiment to verify the efficiency of this algorithm.The experimental results show that the proposed method can extract and recognize the transmission lines with simple calculation and high speed in the movement background. It lays the foundation for achieving all function as an important part of the whole system.