论文部分内容阅读
钢轨三维形貌测量对行车安全有重要的意义,而铁路运营的里程越来越长,为了更准确、高效地复原和测量钢轨的三维形貌,需要对测量图像进行拼接。本文对Harris、SUSAN和SIFT三种图像拼接算法进行了仿真分析和比较,对这三种方法的拼接效果做了评价,仿真结果表明SIFT算法的拼接效果最佳。本文将SIFT算法与FTP相结合,应用于钢轨的三维形貌测量,将拼接后的图像进行FTP复原,得到很好的钢轨三维复原相貌。
The measurement of rail topography is important to driving safety. However, the mileage of railway operation is getting longer and longer. In order to restore and measure the three-dimensional topography of rail accurately and efficiently, the measuring images need to be spliced. In this paper, Harris, SUSAN and SIFT three kinds of image mosaic algorithm are simulated and compared, and the stitching effect of these three methods is evaluated. The simulation results show that the SIFT algorithm has the best stitching effect. In this paper, the SIFT algorithm combined with FTP, applied to the three-dimensional rail topography measurement, the stitched image FTP recovery, get a good three-dimensional reconstruction of rail appearance.