基于友好人机界面的采摘机器人系统构建分析
Construction of Picking Robot System Based on Friendly Human-Machine Interface
-
摘要: 以采摘机器人为研究对象,针对现有采摘机器人采用CCD进行图像识别采集,其数据采集精度和速度均具有一定的局限性的问题,基于友好人机界面构建了采摘机器人的视觉系统。采用激光测距仪作为视觉系统的数据采集装置,综合运用运动控制系统实现激光测距仪的位移变化,选用上位机进行实时数据处理。在视觉系统构建基础上,以果园为目标进行实际检测试验,在视觉系统距离采摘对象不同距离处进行数据采集,绘制出不同距离条件下苹果的轮廓曲线,得出视觉系统对环境当中果树的识别度。试验数据表明:随着距离的增加,视觉系统对苹果轮廓的检测精度逐渐降低,在200mm处对采摘环境的识别度最高。Abstract: In this paper, the picking robot system is studied. The existing picking robot uses CCD for image recognition and acquisition. The accuracy and speed of data acquisition have certain limitations. In order to solve this problem, a visual system of picking robot is constructed based on friendly man-machine interface. Laser rangefinder is used as data acquisition device of visual system, and motion control system is used to realize displacement change of laser rangefinder. Upper computer is selected for real-time data processing. Based on the construction of visual system, the actual detection experiment was carried out with the orchard as the target. Data were collected at different distances from the visual system to the picking object, and the contour curves of apples under different distances were drawn. At the same time, the recognition degree of the visual system to the fruit trees in the environment was obtained. The experimental data show that with the increase of distance, the detection accuracy of the vision system for Apple contour decreases gradually, and the recognition degree of the harvesting environment is the highest at 200 mm.
下载: