Fast Component-Based QR Code Detection in Arbitrarily Acquired Images

Quick Response (QR) codes are a type of 2D barcode that is becoming very popular, with several application possibilities. Since they can encode alphanumeric characters, a rich set of information can be made available through encoded URL addresses. In particular, QR codes could be used to aid visually impaired and blind people to access web based voice information systems and services, and autonomous robots to acquire context-relevant information.

However, in order to be decoded, QR codes need to be properly framed, something that robots, visually impaired and blind people will not be able to do easily without guidance. Therefore, any application that aims assisting robots or visually impaired people must have the capability to detect QR codes and guide them to properly frame the code.

A fast component-based two-stage approach for detecting QR codes in arbitrarily acquired images is proposed in this work. In the first stage, regular components present at three corners of the code are detected, and in the second stage geometrical restrictions among detected components are verified to confirm the presence of a code. Experimental results show a high detection rate, superior to 90 %, at a fast speed compatible with real-time applications.