Font Size: a A A

Research On Positioning Method Of Mobile Robot Based On Vision

Posted on:2020-09-11Degree:MasterType:Thesis
Country:ChinaCandidate:W WangFull Text:PDF
GTID:2428330572484448Subject:Precision instruments and machinery
Abstract/Summary:PDF Full Text Request
The development of robotics has embarked on the fast lane as the country's overall strength has improved,making robots smarter.At present,China has raised the development of robots into a national strategy.The national economy is developing rapidly,and the people's living standards are rising year by year.The demand for mobile robots that can improve people's quality of life is increasing.More and more eyes are on the research and development of indoor mobile robots.However,the application of indoor mobile robots is still Not widely carried out,the technical difficulty and the high price of the sensor are the most fundamental reasons.Therefore,in order to make human life convenient and intelligent,more and more researchers have invested in the field of indoor applications of mobile robots.At present,the biggest problem of indoor mobile robots is how to solve the positioning problem.Only when the robot knows its immediate position and how to reach the target position from the current position can the task be effectively and smoothly completed.It can be seen that autonomous positioning is the basis for robots to complete path planning and navigation planning.There are many ways to achieve positioning.Simple and mature magnetic rails,but there are disadvantages of inconvenient laying;wireless positioning methods include WLAN,Bluetooth,RFID and other distance limitations,the signal will be attenuated with increasing distance,and the accuracy is poor;Ultrasonic waves are easily interfered with by obstacles and other environments;while laser radar and microwave radar are very accurate,they are expensive and not practical for general robots.With the development of technology,visual sensors can meet the general requirements of use in terms of price and accuracy,and can acquire rich environmental information like human eyes,which is beneficial to intelligent control.Therefore,the research on the positioning of mobile robot based on vision is also coming more and more popular.This paper takes mobile robots in indoor environment as the object,mainly studies the use of camera to obtain environmental information to solve the motion information of the robot,and finally completes the positioning estimation of the robot,and proposes an improved algorithm in feature extraction and matching.This article has completed the following tasks through experiments:1)Research on the positioning method of mobile robots.Through a large number of inquiries related literature,in the deeper understanding of the research background and research significance,the related technology of robot positioning is deeply studied,and finally the positioning method of mobile robot based on visual positioning is extended,and monocular vision is extended.The algorithms involved in the field of positioning have been studied theoretically.2)Introduced the imaging principle of monocular camera and theoretically derived the calibration algorithm of Zhang Zhengyou.Experiments were carried out on the indoor lighting environment,and the influence of illumination conditions on the camera calibration parameters was analyzed.The experimental results show that in the indoor scene of this paper,the camera calibration parameters obtained under normal illumination conditions can meet the needs of use.3)The principle of ORB algorithm is elaborated and studied in depth.According to the application scenario of this paper,a feature extraction and matching method of ORB algorithm with adaptive threshold is proposed.Experimental results show that the method improves the effective extraction rate of feature points.With the correct match rate.4)Applying the robot motion model to recover the attitude information of the robot in motion and construct the motion trajectory,and finally realize the visual positioning estimation of the robot.
Keywords/Search Tags:Mobile robot, Visual positioning, Adaptive threshold ORB, Feature extraction and matching
PDF Full Text Request
Related items