Open AccessCITRUS FRUIT QUALITY CLASSIFICATION BASED ON SIZE USING DIGITAL IMAGE PROCESSING
Author(s) -
Ulzii-Orshikh Dorj,
Uranbaigal Dejidbal,
Hiun-Suk Chae,
Lkhagvadorj Batsambuu,
Altanchimeg Badarch,
Shinebayar Dalkhaa
Publication year - 2019
Publication title -
sibirskij vestnik selʹskohozâjstvennoj nauki
Language(s) - English
Resource type - Journals
eISSN - 2658-462X
pISSN - 0370-8799
DOI - 10.26898/0370-8799-2018-5-12
Subject(s) - perimeter , pixel , citrus fruit , mathematics , image processing , horticulture , computer vision , image (mathematics) , computer science , biology , geometry
A new computer vision algorithm for citrus fruit quality classification based on the size of a single tree fruits was developed in this study. The image properties of area, perimeter, and diameter for the citrus fruits were measured by pixels. In order to estimate citrus fruit size in a realistic manner, the ratios of diameter, perimeter and area in pixel values in relation to the actual size of one fruit were determined. The total of 1860 citrus fruits were grouped based on diameter, perimeter, and area in pixels. The results of the grouping of citrus fruits by diameter, perimeter and area were compared with the results of the survey research into citrus fruit size as conducted by the Jeju Citrus Commission. Comparative results reveal that the image of the citrus fruit diameter in pixels demonstrate a more accurate size than the other two pixel values, i.e. perimeter and area.