机器人视觉路径跟踪识别定位优化仿真

Robot Visual Path Following Identification Optimization Simulation

对机器人视觉路径进行跟踪识别定位,可增加机器人在预定路径引导下的障碍物检测准确度。由于对机器人视觉获取的路径图像信息进行高度压缩,使得视觉信息数据存在大量误差。传统的跟踪识别定位方法主要通过对视觉数据分类、特征提取完成定位的,但是对误差小的数据不能准确的与正确数据进行区分,导致跟踪识别定位误差大、不准确的问题。提出采用数据聚类识别的机器人视觉路径跟踪识别定位方法,对数据间的相似度进行计算,根据计算结果对机器人视觉数据聚类成簇,并对簇进行标类,通过超矩形建模算法构建机器人视觉正确数据模型,将标类完成的待定位数据带人视觉正确数据模型中,实现机器人视觉路径进行跟踪识别定位。仿真结果表明,采用所提算法进行机器人视觉路径跟踪识别定位,具有很高的定位精度、所需时间短,定位性能较高。

Following robot visual path identification,can increase the robot in the lead of scheduled path obstacle detection accuracy.Due to the path for robot visual access to highly compressed image information,make visual information data is large error.The traditional tracking identification method mainly through the visual data classification,feature extraction of relocating,but for small error data is not accurate and correct data to distinguish,lead to track recognition positioning error is big,the problem of inaccurate.Put forward based on the data clustering recognition robot visual path tracking identification method,to calculate the degree of similarity between data,according to the calculation results of robot visual data clustering clusters,and to the class of clusters,by rectangular modeling algorithm to build a robot visual data model correctly,will finish the class backlog bits of data into visual data model,right robot visual path tracking identification.Simulation results show that the proposed algorithm is used to identify the robot visual path tracking location,has the very high positioning accuracy,short time required,the positioning performance is higher.