Real Time Algorithm of Wear Detection
This project is about finding how much wear can be detected by off-line image analysis method. The first stage of this project is about image segmentation where the captured image was converted to binary image to ease the process of finding non-zero element in the matrix which is element 1. The proc...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Final Year Project |
| 语言: | English |
| 出版: |
IRC
2016
|
| 主题: | |
| 在线阅读: | http://utpedia.utp.edu.my/17242/1/Dissertation%20Wan%20Abdul%20Azeem%20bin%20Wan%20Rahim%2016289.pdf http://utpedia.utp.edu.my/17242/ |
| 标签: |
添加标签
没有标签, 成为第一个标记此记录!
|
| 总结: | This project is about finding how much wear can be detected by off-line image analysis method. The first stage of this project is about image segmentation where the captured image was converted to binary image to ease the process of finding non-zero element in the matrix which is element 1. The process was repeated using wear edge. Then the result was plotted and the cross correlation between the two edge can be observed. From the result, the cross correlation can detect wear size of 3 pixels (0.14 mm) for that particular image. Further testing was done to illustrate the wear get bigger by using wear size of 63 pixels, 122 pixels and 202 pixels (3 mm, 6mm and 10 mm). The last part is illustration of random and multiple wear to observe the cross correlation pattern. The recommendation that can be made is by using bigger lens to capture the image. |
|---|