Vision-based interaction using mobile devices

• Main Author: Wesley Djingga
• Other Authors: Goh Wooi Boon
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/59244
• Institution: Nanyang Technological University

In this report, computer vision was explored in the form of vision-based interactive game. This game was created to simulate the ability of computer vision to detect simple objects. There were ten standard colors and six standard geometry shapes that the game was able to detect. The game could also identify multiple numbers of color-shape combinations, thus increasing the flexibility and complexity. Furthermore, game design was created and designed to be simple and user friendly, to achieve the best user experience while playing. In order to achieve good accuracy and performance, a number of experiments were performed. Hybrid color space was one of the experiments which had been analyzed. Other important key properties of shape detection were also examined and analyzed, such as brightness, contours, area, elongation and angle. These results were carefully measured to achieve a good game experience. Hence a simple interactive game experience was created.