Vision-Based Identification of Vacant Seats in Public Transport for the Visually Impaired

The accessibility of public spaces for visually impaired individuals is a major concern. One of the challenges faced by blind people in public spaces is the difficulty in locating vacant seats. The presented paper discusses the design, development and evaluation of a computer vision based vacant seat detection system. The system has been developed using various image processing techniques such as BRIEF, Extreme Gradient Boosting, Random Forest. The system identifies features in horizontal lines of the seat to differentiate occupied seats from unoccupied seats. It utilizes various image processing and machine-learning algorithms to improve accuracy and performance. The system has been trained and tested on more than 500 images, which results in quite an impressive accuracy of 81%. The paper highlights the importance of such a system in improving the accessibility and convenience of public spaces for visually impaired individuals.