Research And Application Of Image Retrieval Technology Based On ORB Feature

In recent years,augmented reality technology is one of the research hotspots in many universities and research institutes at home and abroad.Because it applies the virtual information to the real world,it brings new experience to people.At present,the technology has been widely applied in many fields,such as game,teaching,military,medicine and so on,and has been further developing and extending.The development of augmented reality technology cannot be separated from accurate image retrieval technology.With the development of augmented reality technology,the recognized image database is also expanding,which has more urgent requirements for the accuracy,effectiveness and robustness of image retrieval technology.In this paper,the research status of the content based image retrieval method is analyzed,and the image feature extraction and feature coding are studied.The VLAD image retrieval system based on ORB features is realized.On this basis,the design and implementation of an augmented reality travel system is completed.The main research work of this paper is as follows:1.In view of the large and complex image database,how to rapidly and accurately from image retrieved similar problems,we propose VLAD based on ORB feature and cosine distance vector for image retrieval.The experimental results show that compared with the BOW method using SIFT to extract features,the proposed method has obvious advantages in real-time and accuracy of large image retrieval system.2.This paper designs a tourism application software based on augmented reality.The software implements the following functions: when the visitor uses the software to scan the image,the system will match the scanned image to the image in the background image database,retrieve the similar pictures,and then show the relevant introduction about the tourist attractions.A 3D model recognition system is proposed based on two-dimensional image scanning.