基于图像的三维物体测量技术研究

Research on 3D measurement technology based on image

针对非接触性测量问题,提出了利用数码相机对物体进行三维测量的一套实验方案。首先选用技术成熟的Harris算法对被测物体进行特征点提取,其次根据计算机视觉基础理论对数码相机内外部参数进行整体标定,并于标定过程中提出一种利用标准棋盘格图像与传统标定物体相结合的方法。最后,根据该方案进行了实例仿真,与传统相机标定相比提高了精度,简化了求解的复杂性。

To the untouchable measurement question, the feasible reconstruction method is put forward based on the relevant theories of the 3-D measurement. Firstly, several comer detected arithmetic operators are compared based on the existing papers, and the harris operator is selected due to its better stability. Secondly the digital camera internal and external parameter is calibrated as a whole according to the computer vision basic theory, and a new method which use the standard checkerboard image and the traditional camera calibration object is proposed in the process of calibration. Finally, the example is in progress according to this set of schemes simulates, it increases the precision compared with the traditional camera calibration and simplifies the solution complexity.