视觉导航温室机器人路径识别算法与实验  

Path Recognition Algorithm and Experiment of Greenhouse Robot by Visual Navigation

作　　者：张群 宋健[2] 曹高华[1] 袁凤茹 ZHANG Qun;SONG Jian;CAO Gaohua;YUAN Fengru(Collage of Mechanical and Electronic Engineering, Shandong University of Science and Technology, Qingdao 266590, Shandong, China;College of Mechanical and Electrical Engineering, Weifang University, Weifang 261061, Shandong, China)

机构地区：[1]山东科技大学机械电子工程学院,山东青岛266590 [2]潍坊学院机电与车辆工程学院,山东潍坊261061

出　　处：《实验室研究与探索》2018年第5期14-16,41,共4页Research and Exploration In Laboratory

基　　金：国家自然基金资助项目(51505337);山东省自然科学基金资助项目(ZR2014EEP013)

摘　　要：为解决农业机器人温室环境下路径识别实时性较差等问题,以大棚环境下番茄种植为背景,在Lab色彩空间下进行处理,将番茄植株从土壤、地膜等复杂背景下识别出来,通过形态学运算对图像进行降噪处理,利用分水岭算法进行图像分割进而转化为二值图像,根据植物位置获得导航离散点,进而通过Hough变换得到导航路径。为此,设计了一种基于机器视觉的温室机器人,并进行实时图像处理实验。实验表明,该方法在温室环境下能够大幅降低自然光和温室薄膜对图像分割的影响,道路信息提取正确率达到95.7%,基于机器人对连续动态图像分析,该路径识别方法具有较好的适应性和实用性。In order to solve the problem of poor real-time recognition of path in agricultural robot environment. The tomato plants were identified from the complex background of soil and plastic film,and the images were denoised by morphological operation. The image was segmented by the watershed algorithm and transferred into a binary image.According to the location of the plant,navigation discrete points were obtained,and then through the Hough transform the navigation path was established. Then,a greenhouse robot based on machine vision was designed and real-time image processing experiment was carried out. The experimental results show that this method can significantly reduce the influence of natural light and greenhouse film on the image segmentation in the greenhouse environment,and the correct rate of road information extraction is 95. 7%. Based on the continuous motion image analysis of the robot,the path recognition method has good adaptability and practicality.

关 键 词：温室机器人 图像处理 分水岭算法 路径规划 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]