视觉测量系统标定过程中的高精度图像特征提取

An high accuracy algorithm of image features detection for the calibration of visual measurement system

在便携式应变测量系统中,相机标定过程是必须而且关键的一个步骤,针对系统中使用的立体标定模板,建立了一种获取标定图像特征的高精度提取算法。首先经过图像预处理和连通性分析等步骤,对标定特征进行初始的分割和定位,然后采用经典的直线检测方法确定直线特征,计算直线之间的交点做为标定特征点,采用圆形标记编码确定标定平面位置,建立标定特征点图像坐标和空间坐标之间的对应关系。对模拟图像和真实图像进行了处理,通过对大量实测图像进行测试标定实验,结果表明本文方法性能稳定可靠,能够满足测量的需要。

Camera calibration is a necessary and key step before applying machine vision to measure- ments. Aiming at stereo calibration sample with line grids, a high accuracy method for detecting image features is presented. The image of calibration template is obtained with image processing and pixel connection analysis, the initial features image is segmentation and location. Grid lines of the calibration sample image are distinguished using Hough transformation, and locates the circle labels on template plane using image technique, and accurate image features are detected. The method presented is applied to images with simulated noise and real images. The experiment results shows that the algorithm is reliable and stable and it can be applied in real scene.