Research On Automatic Detection Of Target And Self-localization Technology With Single-view Based On Image Process

With the development of the computer technology, automatic detection of target based on image processing is more and more widely used in the fields of industry. It is also the hot spot issue of the mobile robot self-localization. Therefore, the article does some research on automatic detection and self-localization of mobile robots as follows.Firstly, in the filed of digital film processing, an automatic detection algorithm based on the fuzzy theory is presented, which is used for automatic defect detection and recognition. This paper also studies the preprocessing of film images and proposes an algorithm for extracting information from a welding seam. Based on this algorithm an intelligent management system in dealing with the target detection is set up. The system can automatically detect defects of the digital film using the method of fuzzy pattern recognition, and also provide an efficient intelligent management for the detection results, films, RT reports and staff etc. At the same time, the interface of the data import/export is also supplied in the system. Moreover, a backstage database is set up in the system to store the above-mentioned data for future use, which can provide statistical data for welding quality management and be served as a reference to the engineers.Secondly, the paper deals with the working principle of the monocular mobile robot self-localization based on the target detection of image. In the beginning, this paper argues an artificial landmark detection algorithm and obtains an identifying code for each landmark according to the image processing. Then this paper studies the solution for the P4P problem, which is used in the mobile robot's position calculation. At last, the experiment results and algorithm errors are analyzed.Experimental findings show that the proposed method can not only identify detects effectively, but also obtain the type and location of defects accurately. Facts have proved that the robot localization system based on monocular vision and artificial landmark can be applied to autonomous mobile robots successfully. The system designed by this paper can be widely used in the realm of industry and also can greatly save labor force and increase production efficiency.