Up to journal
for Research & Developement
Journal of Materials Science: Materials in Electronics
Impact Factor
0/000
Available
2000 - 2020
Volumes
8
Issues
25
Articles
218
Open Access
66


Journal of Researches in Mechanics of Agricultural Machinery
Design of Tangerine Sorting Algorithm based on Color Using Image Processing


 submission: 11/10/2019 | acception: 15/02/2020 | publication: 10/06/2020

DOI 

Authors
Ali Nejat Lorestani1*, Kayvan Yazdanpanah2, Sajad Sabzi3

1-Razi University،ali.lorestani@gmail.com

2-Razi University،sheno135@yahoo.com

3-University of Mohaghegh Ardabili،sajadsabzi2@gmail.com



238 Downloads

Abstract

Many citrus fruits, such as tangerines in the country, are abandoned in orchards every year due to their low prices, causing great damage to the gardeners. On the other hand, the supply of homogeny product in terms of the amount ripeness, to the market, will be welcomed by the consumer and will stimulate the market for product sales. Tangerine has different varieties that in this research, Unshiu variety of Tangerine has been studied and image processing and artificial neural networks (ANN) were used for classification of tangerines into three degrees of ripe, half - ripe and unripe. ۱۲۰ Tangerine fruit samples that graded by Expert Grading person, were selected And then their images were obtained using an imaging system. After taking pictures of tangerines, pre - processing operations, Segmentation operations and Classification of images were done and Images' attributes transferred to different color channels Specified such as RGB, HSV, YCbCr and CMY and the statistical characteristics of the images were extracted. Average of second component of YCbCr color space, standard deviation of third component of YCbCr color space, standard deviation of third component of CMY color space, average of first component of HSV color space, average of third component of HSV color space and standard deviation of second component of HSV color space were selected as input of proposed system. In total ۶ × ۴ Characteristics were extracted from each sample. Overall, the percentage of the correct classification of the three classes was ۹۷.۲۲ %. The results of this study indicated that the proposed system is capable to detect tangerines ripe, half - ripe and unripe carefully fit it had. Therefore, it can be concluded that modern methods such as image processing and artificial intelligence are used to classify tangerines.




Keywords

image processing  artificial neural network  sorting  color  classification  Tangerine 



Download fulltext PDF


Open Access

Download

Citation





References





Announcements  

  •  No announces available
Indexes  

 

no item exists