基于局域最大亮度的靶状激光图像边缘检测

Edge detection for laser image of target-shape based on local maximum intensity

针对工程应用中靶状激光图像带有较强噪声干扰和要求图像处理速度快的特点,本文提出基于局域最大亮度的靶状激光图像边缘检测方法。该方法根据靶状激光图像由多个明暗相间的圆环组成的特征,采用一个大小适当的圆形检测模板,让模板沿着亮环按一定规则在一定范围内移动,计算模板内各像素点的亮度和,使亮度和为最大值的模板的中心点为这个区域的边缘点。用最小二乘法将由此方法获得的边缘点拟合定中。用摄像机(640像素×480像素,光靶分辨率为0.0504mm/像素)距光源10m处拍取50幅图像,处理后的定中标准差为0.0237mm。实验和应用结果表明,此方法速度快,精度高且抗噪声性能好。

Since the laser image of target-shape is strongly noised and the image processing must be fast in industrial application, an edge detection method for laser image of target-shape based on local maximum intensity is proposed Based on this type of image composed of several light-shade alternating rings, we used a round detecting mask of suitable size and moved the mask along a light ring in a specified way and in a certain range. Then we calculated the sum of intensity of all pixels in the mask and chose the center of the mask with maximum sum as an edge point in the local area. All the edge points gotten by this method were fitted to obtain edge center with least square method. For 50 samples from a camera （640×480 pixels and 0.0504mm/pixel as resolution of the target） at distance of 10m, image processing was carried out. Results show that the standard deviation is 0.0237mm. Experiment and application instance demonstrate that this method has a fast processing rate, high accuracy and robustness to noise