基于立体视觉的农机热锻件测量系统设计

为解决农机热态锻件的在线尺寸测量问题,提出了一种基于计算机双目立体视觉的解决方案。针对热锻件的图像获取、特征提取、图像匹配和三维重建等问题展开研究,分别采用光谱选择性图像采集方案、形态学角点和轮廓提取方案,基于标定立体校正的快速匹配和重建方案,构建了一套在线测量系统。为提高系统的实时性,采用Open CV计算机视觉库函数实现相应算法。以一个被加热至1 000℃圆柱体锻件为试验对象,测量结果直径相对误差为2.20%,高度相对误差为1.81%,用时87s。试验结果表明:所提出的农机高温热锻件几何参数测量方法能够满足锻造生产现场对尺寸测量实时、精确、高效的要求。

The Measurement System Design of the Agricultural Machinery Hot Forging Parts Based on the Stereoscopic Vision

Aiming to solve the online measurement problem exists in hot forging process of agricultural machinery compo-nents, a solution was explored base on the stereo vision of computer in this study.The image acquisition, feature extrac-tion and 3D reconstruction of hot forging parts were studied.A set of real-time measurement system were constructed with spectra selective image acquisition, morphology corner detection and contour extraction,rapid matching and 3D re-construction based on stereo rectification.In order to enhance the real-time function of system, the relevant algorithm was realized by using functions of OpenCV.The experiment, which was based on cylinder forging ( heated to 1000℃, indicated that the relative errors of measurement diameter and height were 2 .20%and 1 .81%by using 87 s.The experi-ment results indicated that the geometric measurement method of agricultural machinery hot forging parts could satisfy the requirements of real time, precision and high efficiency of dimension measurement on production spot.