摘要
当前智能车、智能机器人技术发展迅速,基于简易视觉的障碍物检测方法愈加重要。为此,结合部分场景障碍物检测进行研究,建立了基于拟合和模糊推理的单目测距模型。首先,选择某些典型物体作为标准可识别物,并以简单函数的拟合曲线近似其图像与距离之间的数量关系;然后,对测距场景进行模糊识别,锁定特定的标识物,以简化测距模型匹配的计算量;最后,为提高测算精度,对障碍物与标准可识别物的贴近度进行计算,选用贴近度最高值来测算障碍物与镜头之间的距离。实验结果表明:该方法具有一定的精确性,可以满足智能机器人测距的实时性和准确性要求。
With the rapid development of intelligent cars and robots,the detection method of obstacles based on simple visual becomes more and more important.In this paper,a single-camera vision distance model based on fitting and fuzzy inference is proposed by studying the obstacle detection in some scene.First,some typical objects are selected as the standard identifiable objects,and the quantitative relationship between the image and the distance is approximated by a fitting curve of a simple function.Second,the fuzzy identification of the ranging scene and the identification of the specific identifier are used to simplify the calculation of the matching of the ranging model.Finally,in order to improve the accuracy of measurement,the degree of closeness between the obstacle and the standard identifiable object is calculated,and the closest identifiable object is used to measure the distance between the obstacle and the lens.The actual measurement results show that the method is real-time and can meet the accuracy requirements of intelligent robot ranging.
作者
王志刚
赵海良
王星
WANG Zhigang;ZHAO Hailiang;WANG Xing(School of Mathematics,Southwest Jiaotong University,Chengdu 611756,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2020年第1期58-63,共6页
Journal of Chongqing University of Technology：Natural Science
基金
国家自然科学基金项目(61473239)
关键词
机器人测距
图像识别
模糊推理
intelligent robot measuring distance
image recognition
fuzzy inference
