Journal of Engineering, Project, and Production Management, 2026, 16(3), 2025-232
Design of an Indoor Visual Positioning System Based on Incremental Learning and Multi-Feature Fusion
1
Associate Professor, School of Art and Design, Fuzhou University of
International Studies and Trade, Fuzhou, 350202, China.
2
Instructor, School of Urban Design, Shanghai Art and Design Academy,
Shanghai, 201800 China, E-mail: sunzpingzp@outlook.com (corresponding
author).
Project Management
Received October 15, 2025; revised December 8, 2025; accepted December 29, 2025
Available online April 8, 2026
Abstract: Indoor visual positioning in complex indoor spaces requires reliable classification and detection under changing lighting, occlusion, and multi-category interference. To enhance recognition accuracy and real-time performance, a visual orientation system, integrating multi-feature fusion, incremental learning, and a regional proposal detection mechanism, is developed. Deep convolutional, gradient structural features, and texture descriptors are fused to enhance scenario representation, while incremental learning enables category expansion without degrading existing recognition ability. The regional proposal detection module improves target localization and boundary fitting in cluttered indoor layouts. Experimental results on the SUN RGB-D dataset show that the classification model achieves an accuracy of 0.92, a precision of 0.91, a recall of 0.90, an F1-score of 0.91, and a frame rate of 48 frames per second. The detection model achieves an accuracy of 0.94, a precision of 0.92, an Intersection over Union (IoU) score of 0.83, and a processing speed of 38 frames per second. These findings demonstrate that the proposed model achieves a strong balance between recognition performance and operational efficiency, offering practical support for indoor navigation and spatial orientation in dynamic environments.
Keywords: Incremental learning, multi-feature fusion, support vector machine, regional proposal network, visual communication design.
Copyright © Journal of Engineering, Project, and Production Management (EPPM-Journal).
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Requests for reprints and permissions at eppm.journal@gmail.com.
Citation: Xie, Y. and Sun, Z. (2026). Design of an Indoor Visual Positioning System Based on Incremental Learning and Multi-Feature Fusion. Journal of Engineering, Project, and Production Management, 16(3), 2025-232.
