Situation-based Indoor Wayfinding System for the Visually Impaired
Eunjeong Ko, Jin Sun Ju, Eun Yi Kim · 2011 · The Proceedings of the 13th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/2049536.2049545
Summary
This paper presents an indoor wayfinding system for visually impaired people that uses situation-aware color codes affixed to building walls to guide users to destinations in unfamiliar environments. The key innovation is a situation-based approach: rather than providing all environmental information at once, the system first determines the user's current situation (whether they are at a door, in a corridor, in a hall, or at a junction) and then selectively detects and presents only the relevant navigational information. The system uses three types of color codes designed to mimic real-world visual clues: color-barcodes (encoding room numbers using 8 color patches in quinary encoding), pictograms (10 standard building icons for toilets, exits, elevators, etc.), and guide codes (combinations of location codes with directional arrows). These codes are physically affixed to walls near doors, elevators, and corridors. The system runs on an iPhone 4G and operates in three stages: situation awareness using SURF (Speeded-Up Robust Feature) descriptors matched against a vocabulary tree of 200 template images; hierarchical color code detection and recognition using color segmentation and template matching; and auditory output via text-to-speech.
Key findings
The system achieved 95% accuracy in real-time field tests with four visually impaired participants navigating unfamiliar indoor environments to multiple destinations. Situation awareness accuracy exceeded 88% for door, corridor, and hall classifications. Color codes were detectable at distances up to 2.5m with nearly 100% accuracy at viewing angles within ±20°, with zero false positive rate across all distances tested. The system outperformed a comparable QR code-based approach in several ways: QR codes were only detectable within 1m, while the proposed color codes worked up to 2.5m; and beyond 1m the color codes maintained a 30° viewing angle tolerance compared to QR codes becoming undetectable. Average processing time was 227ms per frame, enabling real-time operation at up to 10 frames per second. In field tests across two test maps with three destination goals each, all participants followed near-optimal routes with an average lateral deviation of only 0.78m from the path taken by sighted users, and average decision error rates of 5.6% and 5.0% for the two maps.
Relevance
This research demonstrates a practical vision-based alternative to GPS for indoor navigation, addressing a critical gap since GPS signals are unavailable indoors and technologies like RFID and WiFi provide insufficient spatial precision for visually impaired users. The situation-based approach — delivering different information depending on whether the user is at a door, corridor, or junction — reflects sound accessibility design by reducing cognitive load and providing contextually relevant guidance. For practitioners, the comparison with QR codes is particularly useful: while QR codes are ubiquitous, their short detection range (1m) makes them impractical for wayfinding where users need to detect markers from a distance. The color code system's ability to work at 2.5m and under varying lighting conditions makes it more realistic for deployment in public buildings like hospitals, shopping malls, and schools. The main limitation is the requirement to physically install color codes in buildings, though the codes are designed to complement existing signage rather than replace it.
Tags: blindness and low vision · indoor navigation · wayfinding · computer vision · assistive technology · mobile accessibility · navigation