CAJ | 학술논문

为了提高采摘机器人的定位速度,对机器人的机器视觉系统进行了改进,设计了一种基于聚类算法和视频对象提取技术的快速定位机器人. 该机器人视频对象图像提取过程中,在完成图像进行滤波后,引入了 Lab彩色空间聚类算法,有效地降低了图像的色彩数和噪声,实现了图像对象的量化处理,大大提高了果实定位和采摘的效率. 为了验证设计的快速定位采摘机器人的可靠性,对机器人的采摘性能进行了测试,测试项目主要包括图像处理和果实定位. 通过测试发现:快速定位机器人可以有效地实现图像聚类中心的提取,并对聚类中心进行编码,每次定位用时少、定位速度高且果实采摘的准确性累计概率较高,符合高精度、高效率果实采摘机器人的设计需求.
위료제고채적궤기인적정위속도,대궤기인적궤기시각계통진행료개진,설계료일충기우취류산법화시빈대상제취기술적쾌속정위궤기인. 해궤기인시빈대상도상제취과정중,재완성도상진행려파후,인입료 Lab채색공간취류산법,유효지강저료도상적색채수화조성,실현료도상대상적양화처리,대대제고료과실정위화채적적효솔. 위료험증설계적쾌속정위채적궤기인적가고성,대궤기인적채적성능진행료측시,측시항목주요포괄도상처리화과실정위. 통과측시발현:쾌속정위궤기인가이유효지실현도상취류중심적제취,병대취류중심진행편마,매차정위용시소、정위속도고차과실채적적준학성루계개솔교고,부합고정도、고효솔과실채적궤기인적설계수구.
In order to improve the picking robot localization speed, machine vision system of the robot was improved to design a clustering algorithm and a video object extraction technique based on the rapid positioning robot.After the com-pletion of image filtering,the robot image video object extraction process was introduced into the lab color space, and it used clustering algorithm to reduce the image color and noise, which can realize the quantification of object image pro-cessing, thereby greatly improved fruit location and picking efficiency.In order to verify the reliability of the design of the fast positioning picking robot, the robot's picking performance is tested,which included image processing and fruit lo-cation.It was found by test that the rapid positioning robot can effectively realize the extraction of image clustering cen -ter, and code of the cluster center, fewer in each location, high positioning speed, fruit picking accuracy of cumulative probability is higher,which meet the design requirements of the precision and high efficiency fruit picking machine.