Please use this identifier to cite or link to this item:
Title: Indoor Navigation by Image Recognition
Authors: CHOI, IO TENG (蔡耀霆)
Department: Department of Computer and Information Science
Faculty: Faculty of Science and Technology
Issue Date: 2017
Citation: CHOI, I. T. (2017). Indoor Navigation by Image Recognition (Outstanding Academic Papers by Students (OAPS)). Retrieved from University of Macau, Outstanding Academic Papers by Students Repository.
Abstract: An indoor navigation system using image recognition method is proposed in this paper. The structure of the proposed method can be classified into two main parts: indoor positioning and navigation. We propose to use image recognition to achieve indoor positioning. Images of indoor environment are first collected and features are extracted to form a feature collection. Then, getting real-time images from the user, and perform feature extraction. Through the comparison of the result and the collection, the current location of the user can be recognized. To achieve real-time computing effect, DARTs features which have a faster performance than SURF and SIFT are used for recognition process. REVV algorithm is then used to generate global signature for the local features to have a faster performance in storing and searching. Once the position is confirmed, navigation will start. On the navigation process, Dijkstra ‘s algorithm is used to perform the shortest path search on the indoor map graph and return a route to a certain destination. Based on the resulted route, user’s position will be kept tracking to confirm that user is within the desired path. Otherwise, path calculation is performed again. Our proposed system is implemented on an Android platform for testing. The experimental results indicate that the error rate of our proposed method is low and the performance of the system is satisfied. Also, indoor navigation by image recognition can overcome the disadvantages and the constrain of other indoor navigation method.
Instructor: Prof. PUN, CHI MAN
Programme: Bachelor of Science in Computer Science
Appears in Collections:FST OAPS 2017

Files in This Item:
File Description SizeFormat 
OAPS_2017_FST_014.pdf3.04 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.