A Study of the Detection for Betel Nut with Machine Vision
| Autor: | Chih-Hong Kuo, 郭志宏 |
|---|---|
| Rok vydání: | 2008 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 97 The manufacturers of betel nuts commonly use machines to separate betel nuts according to their sizes, but large amount of laboring is still required to maintain the quality of the product. In order to cope with the market growth of betel nuts, manufacturers has been looking for ways to advance to an automation system to cut down the cost of the process and to improve the quality control. This project is to study the application of the visual function of a machine to separate betel nuts into group base on their physical appearance. The methods used for this application includes betel nut characteristic value computation rule, betel nut spot examination computation rule, and betel nut apparent and size classification principle. This project has studied the following fields: 1.Establishment of calculation of betel nut characteristic value: (1) Use image process technology to establish estimated betel nut geometry characteristic value. (2) Design the betel nut geometry characteristic examination program. 2.Establishment of betel nut spot examination principle: (1) Use image process technology to establish betel nut spot examination principle. (2) Design the betel nut spot examination program. 3.Establishment of betel nut classification principle. (1) Use image process technology to establish the betel nut classification examination principle. (2) Design the betel nut apparent classification examination program. (3) Design the betel nut size classification examination program. This project study used the image process technology. Estimated characteristic values, such as the length, width, and shape proportion of betel nuts were obtained by computation after the images undergo logic and binaryzation operations. Classification methods to investigate the spot area, the total surface area, and to group the betel nuts based on their outward appearance and size has been established. The results of this project is obtained by studying 400 betel nut samples, and the accuracy of the results is 92 %. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|