论文部分内容阅读
为了克服铁路建设中人工检测的局限,在满足测量精度的条件下,利用图像处理方法,对基桩进行在线识别。由CCD采集基桩的图像,对基桩图像先进行直方图均衡,再用模板匹配法对其进行识别。在实验中,对用直方图均衡处理和其他图像预处理的方法进行了比较。结果表明,用模板匹配法能将经直方图均衡过的图像中的轨道基桩有效地识别出来,且识别精度高、速度快,可以满足铁轨在日常铺设时的测量要求。
In order to overcome the limitations of manual testing in railway construction, the paper carries out on-line identification of pile foundation by using image processing method under the condition of meeting the measurement accuracy. The image of the pile is collected by the CCD, the histogram of the pile is balanced first, and then the template matching method is used to identify the pile. In the experiment, the methods of histogram equalization and other image preprocessing were compared. The results show that the method of template matching can effectively identify the track piles in the histogram-equalized images with high recognition accuracy and high speed, which can meet the measurement requirements of daily laying of tracks.