Abstract
The purpose of this paper is to provide reasonable recommendation and removal of inappropriate information for SLAM (Simultaneous Localization and Mapping) technology based on feature method. The methodology is to propose a semantic recognition of environment objects in the natural scene through object detection, which is a kind of bag of word method in SLAM problem between the key frames and object level, the method of establishing key frames, and the relationship between the target object levels, through the practical significance of the target object level to judge the merits of the target object level information, and then combined with key frames in the visual SLAM relations with relevant information, so as to get object level targets in each key frame and the relationship between the relevant information, so as to achieve through the object level semantic information to judge the merits of the key frames and screening, as well as to the key frames to judge the merits of the relevant information and screening purpose. The finding of the study is the above method can retain the information of high reliability and good stability for visual SLAM and process the key frames with poor reliability or low stability and the information related to key frames.
Highlights
For visual Simultaneous Localization and Mapping (SLAM), in order to ensure the stability of positioning and mapping, key frames with good stability and relevant information should be retained as far as possible [1].e relevant information here refers to the information used for mapping and map-ping correlation calculation in visual SLAM, since the front-end of visual SLAM is divided into direct front-end and indirect front-end
YOLO, this paper provides semantic information and adds a filtering wheel combining semantic information before ORB- SLAM2 original map points to filter the correct markers, filtering the features and placing key frames on the map points relative to the dynamic objects after filtering the step to calculate each object corresponding to the four formulas of key frames, each product of the probability of the corresponding N object category, and comparing with the corresponding threshold, in order to exceed the threshold map points associated with this object, optimize the total view of the key frames
While maintaining ORB-SLAM2’s original screening process, in order to ensure a good relationship between the signposts and key frames, the following two conditions are required for good signposts to be screened: (1) More than 25% frames can be observed in theory
Summary
For visual SLAM, in order to ensure the stability of positioning and mapping, key frames with good stability and relevant information should be retained as far as possible [1]. E relevant information here refers to the information used for mapping and map-ping correlation calculation in visual SLAM, since the front-end of visual SLAM (visual odometry) is divided into direct front-end and indirect front-end. For the former, this relevant information, namely, feature points, such as SIFT, SURF, and ORB, are effective feature points to be extracted. For indirect visual SLAM, the relevant information is the bright ness represented by a single pixel In these SLAM methods, it is assumed that all motion estimation is carried out under a relatively ideal premise; that is, the information has good invariability with the change of time and space. E method proposed in this paper hopes to carry out semantic understanding of relevant information in the image through the object detection method and filter the key frames and relevant information obtained to eliminate highly dynamic objects and unstable objects in semantic concept, so as to improve the positioning robustness of SLAM system
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
