The Application Of Bionic Algorithm In Mobile Robot Visual Location Research

In recent years,with the progress of the modernization of industry and the needs of the people’s life for convenient and comfortable,the robot began to become the focus of the study hotspot.Using robots can make people do project construction in place where against manual operation,such as the specially heavy pollution,corrosion environment.Afterwards,as the world’s population decline,the cost of human labor has greatly promoted,in this situation,the robot application is particularly important in the industrial process,it can saves a lot of manual labor and reduce the production cost of enterprises.The application of artificial intelligence based on internet of things in recent years,makes people’s lives convenient and quick furthermore.Because the robot has the advantages of the above,so it can becames the world industrial development trend in the future.In the process of robot research,machine vision as one of the most important part of robot,has a large number of researchers.Robot vision systems including monocular,binocular and many eyes vision system.They each have advantages and disadvantages,of which, binocular vision system is the mainstream now,more eyes vision system because of its complexity structure and higher cost, so we do not use more eyes vision system generally. Monocular vision system has simple structure, easy installation, small volume of the whole system occupies,and can realize the task which binocular vision system can complete after converting to the binocular vision use a certain method,so this article use monocular camera convet to binocular vision to realize the robot vision orientation.In the process of visual location,we use improved SURF algorithm,SURF algorithm is based on SIFT algorithm,and it is the improved SIFT.It not only has the same properties with the SIFT algorithm, but also it has more speed in the whole processing.Because unimproved SURF algorithms easily occur false match,so SURF algorithm is improved at the end of the paperjoined RANSAC algorithm to eliminate false matching point.Experimental results show that improved SURF algorithm can eliminate false matching points effectively,the accuracy of visual location was improved greatly.