除草机器人路标识别多模式匹配算法研究——面向中英文混合环境

随着农业自动化水平的不断提高,除草机器人被逐步地应用到了农田生成作业过程中,机器视觉系统是除草机器人自主导航和杂草作物识别的核心部件,其性能的好坏直接影响作业效率和作业质量。为了提高除草机器人的导航效率,提出了一种基于中英混合环境的多模式匹配算法,将该算法应用到了除草机器人的英文和中文导航路标的模式匹配上,并采用相关图像处理算法对路标进行了增强处理,最后对不同环境下导航的可行性进行了验证。结果表明:采用中英文混合环境的多模式匹配算法和图像增强处理功能后,除草机器人机器视觉系统的路标成功识别率较高,导航准确率也较高,从而验证了方案的可行性和可靠性。

Research on Multi Pattern Matching Algorithm for Road Sign Recognition of Weeding Robot Facing to Mixed Environment for Chinese and English

With the continuous improvement of the level of agricultural automation,weeding robot are being used increasingly to the farmland generated during the operation,the machine vision system is a core component of weeding robot navigation and weed crop identification,its performance directly affects the quality of the work efficiency and work.In order to improve the efficiency of weeding robot navigation,presents a multi pattern matching algorithm based English mixed environment,and applies the algorithm to match the English and Chinese weeding robot navigation sign pattern,and the related image processing algorithm for the enhancement of signs,finally verified by the feasibility of different environment the navigation.The experimental results show that after the multi-mode matching algorithm and image enhancement processing function of Chinese and English mixed environment,the roadmap recognition accuracy of the weeding robot vision system is high,and the accurate navigation rate is high,which verifies the feasibility and reliability of the scheme.