提出了一种目标点全视场自动定位方法,该方法利用回光反射标志点的强反射特性和目标点特征,采用粗精两级定位自动获取目标点中心.实际操作中只需预先设定相关的阈值参量,可快速获取图像中的所有目标点的图像坐标,极大地提高了视觉测量的效率和自动化程度.对于3008×2000pixel大小、包含约140个目标点的RGB图像,利用该方法定位所有目标点的用时为14.9”,定位准确度为1/10pixel.
A method which can be locate all the targets in one image automatically was presented. In practical application, some threshold parameters were presetted according to the high-reflective characteristic and the image features of the retro-reflective targets,and then coarse and fine location steps were performed in turn. Using the method, the efficiency and the degree of automation in practical measurement can be largely improved. For one RGB image with 3 008 × 2 000 size and about 140 targets, the locating time for all the targets is 14.9",and the location precision is 1/10 pixel.