A comparative study of edge detection techniques for AI-based image recognition

• Main Author: Wang, Di
• Other Authors: Meng-Hiot Lim
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: https://hdl.handle.net/10356/141301
• Institution: Nanyang Technological University

Artificial Intelligence (AI) is currently the hottest technology in computer science. It is composed of different fields, such as machine learning, computer vision, etc. In general, a major goal of artificial intelligence research is to enable machines to perform complex tasks that normally require human intelligence to complete. The development of these machines has benefited people's lives. One of the most important steps in the robot's execution of the task is target recognition. In this project, basic image recognition for target recognition is studied. This topic has studied image recognition through two aspects. The first aspect is edge extraction, and the second aspect is training neural networks for image recognition. In this project, different edge detection technologies are studied. We use digital image processing methods to compare several edge detection technologies and observe the effects of different edge shapes, noise and other external factors on their detection results. Finally find the most suitable edge extraction technology for the subject. At the same time, we used neural networks to identify the targets in the image and compared the impact of different sizes of training data on the testing results. Key words: edge extraction, neural network, image recognition