Public signs detection in subway station images

It's difficult to find location in indoor environments such as subway station and airport passenger terminal. However, the detection and recognition of public sign make it possible to find out the place to go. Proposed system can help visually impaired persons or mobile robots in unfamiliar places. We propose the detection method of public signs in natural images. Especially, we dealt with several signs in subway station such as numbers, arrows, ticket, elevator, escalator, phone, toilet, and arrows giving the direction of those places.