Image detection program method C improvement

• Main Author: Zhang, Meixia
• Other Authors: Chia Liang Tien
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/70257
• Institution: Nanyang Technological University

Method C is the new, fast but simple program which written by the project group in Rolls – Royce@NTU Corporate Lab. The method uses the knowledge of Discrete Fourier Transform (DFT), Gaussian filtering and selecting threshold value to detect the topographic and crystallographic defects. But it has some limitations to detect the crystallographic defects. This project is supported by the project group of Rolls – Royce@NTU Corporate Lab. The programming language used in this project is MatLab. The purpose of this project is to improve Method C to deal with crystallographic defects. The basic processes for the project are analyzing the image sets, learning Method C, selecting suitable parameters for Method C, and 3 experiments to select the input image for detection. Finally, the improved Method C can detect the crystallographic defects – HAB and Rx. For fu rther improvement, can classify the types of detected noised and remove them for increasing accuracy.