基于机器视觉的吊具位姿检测系统

摘要

The paper presents a kind of monocular machine vision based spreader detection system (SDS) to detect spreader position. The SDS can measure the spreader position in real time during operating the containers. Firstly, choose a datum mark and install a near-infrared camera on the trolley, and assemble a structuring light source made up of three lamps on spreader. Image processing algorithms are employed to recognize the coordinates of lamps in picture derived from the camera. According to the lamps coordinates, camera calibration parameters and coordinate relationship between the datum mark and the camera, the spreader position and orientation can be worked out. The system has all-weather, no touch, quick and accurate detection, and low cost quality, and can provide valid and reliable feedback data for spreader control system.%提出1种基于单目机器视觉的吊具空间位姿的传感检测系统,在装卸箱过程中实时检测吊具位姿.首先在起重机小车上选取基准点,并安装近红外相机,在吊具上安装由3个发射灯组成的结构光源.用单目相机拍摄吊具上的结构光源,并在暗图像中通过图像识别算法识别出光源在相机中的坐标.然后根据相机标定参数和相机、基准点坐标关系等得到吊具在基准点坐标中的位置和姿态.相较于现有的设备,此检测系统具有成本低、全天候、快速准确等特点,其能为吊具控制系统提供了可靠的反馈数据.