Having OpenCV visualization toolkit based to develop a three-dimensional measurement system for logistics cartons was implemented,including having binocular vision checkerboard used to calibrate the camera and determine its intrinsic and extrinsic parameters;making use of Canny operator to measure edge of the image so as to obtain a proper edge feature and finally,under the condition of epipolar geometry and line constraint,having character points interpolated to complete stereo matching of the image and calculate three-dimensional coordinate of the world coordinate system.The experimental results show that,the system's measurement accuracy can be controlled less than 2mm and satisfy the requirements of classification accuracy.%基于OpenCV视觉函数库,开发了一个物流纸箱三维尺寸测量系统.首先,利用双目视觉棋盘格点进行摄像机标定,求出摄像机的内外参数;然后,采用Canny算子对图像进行边缘处理,得到合适的边缘特征;最后,在对极几何和极线约束的条件下,对强特征点进行插值处理,完成图像的立体匹配,并计算出世界坐标系下的三维坐标.实验结果表明:该系统的物流纸箱三维尺寸测量精度可以控制在2mm以下,满足分类精度要求.
展开▼
