An improved visual SLAM algorithm based on wide-angle camera correction model

Visual SLAM has gained a lot of interest recently, since various vision sensors are important components of mobile robot system. In order to resolve the problem of distortion in wide-angle images and improve the accuracy of feature points matching, an improved visual SLAM algorithm based on the wide-angle camera correction model is proposed. Firstly, the images are divided into blocks and the details of some selected blocks are enhanced. Secondly, the feature points of the blocks are extracted, and by using the feature points of the blocks in the images, the feature points on the images are extracted. Finally, the points on the incident light through the normalized plane are selected as the template points, and the relationship between the template points and the images are established through the wide-angle camera model, and the pixel coordinates of the template points in the images and the descriptors are calculated. Moreover, a series of experiments are conducted in this paper using the TUM dataset. The experimental results show that the trajectory error and translation error measured by the improved visual SLAM algorithm with a wide-angle camera are reduced, compared with the current VINS-MONO system and ORB-SLAM3 system.