Development of a general-purpose vision system for robots and vehicles

• Main Author: Aliyev, Tural
• Other Authors: Xie Ming
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Mechanical engineering
• Online Access: http://hdl.handle.net/10356/78393
• Institution: Nanyang Technological University

With remarkable increase in computing power and the demand for affordable robotic systems increasing exponentially, there is a huge need to fill in the gap with an affordable solution for a robot's interaction to the human world through vision. A robotic vision system gives the robot an advantage to understand its environment and perform simple tasks associated with it. Existing vision systems cater to only specific needs of the user despite of its broad application fields. Thus, the principal objective of this project is to develop a commercially viable general-purpose vision system for robots and autonomous vehicles. The project consists of two significant parts which are hardware and software. Firstly, the project integrates a vision system prototype comprised of a Raspberry Pi microcontroller, camera module and mobile robot car in terms of hardware. Secondly, software library functions and algorithms behind the library functions are developed using programming language Python, open source library OpenCV and various recognition models. Developed algorithms include colour detection, geometric shape detection, object detection, face recognition and light detection. The vision system with its general-purpose libraries allows developers to build their products according to their creativity.