2014년 컴퓨터종합학술대회
Current Result Document :
| 한글제목(Korean Title) |
Automated Quality Analysis for Classification of Rice Grains |
| 영문제목(English Title) |
Automated Quality Analysis for Classification of Rice Grains |
| 저자(Author) |
Joystna Gajanan Anandache
Sung-Hwan Jung
|
| 원문수록처(Citation) |
VOL 41 NO. 01 PP. 0626 ~ 0628 (2014. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
This paper presents an improved method for quality analysis and classifying different types of rice grains. In the grainhandling system, the rice grain type and quality are rapidly inspected by visual inspection and the decision is taken by the food inspector. This decision can be affected by his/her physical condition and surrounding environmental factors. And hence the consumers and the farmers are influenced by this manual inspection activity. We can solve this problem by using image processing techniques. Five different kinds of rice samples images are captured in good intensity of light and are used for tests to check its quality and classify its rice type. This process extracts four geometrical and three color features from rice sample images and are used for analysis. The image processing technique uses probability neural network algorithm to classify type of rice grain and also it counts the number of rice grains present in the given rice sample image. As a result of our test, we got maximum classification accuracy as 91.2%.
|
| 키워드(Keyword) |
machine vision system
grain quality
counting
neural network
matching
recognition
classification
|
| 파일첨부 |
PDF 다운로드
|
