Abstract:
In this Report, a proficient technique of image processing has been used to detect
artificially ripened mangoes. Mango is one of the significant fruit crops across the
world and if we mention Bangladesh, it is the most popular one during summer. Even
the word mango has been originated from manga in Indian subcontinent. But now a
days, this fruit is being ripened artificially with chemicals by some mischievous sellers
unknowingly or sometimes knowingly the hazardous effects of this chemicals on
human health. Among those chemicals that has been used for ripening mango
artificially, calcium carbide is the most used one. Lately, the national agency for food
and drug administration and control (NAFDAC) warned the mass people against
devouring fruits ripened with calcium carbide by saying that those fruit cause heart,
kidney and liver failures when consumed. Hence, consumers must have to be careful
while buying fruits. It is kind of difficult for human eye to find out the fruits that have
been chemically ripened. Here in this research, a method of digital image processing
has been used to aid the detection. This research considers sample images of more than
150 mangoes. Total 24 features have been extracted from this images and then
categorization is done using j48 decision tree, multilayer perceptron and naïve bayes.
Also a comparison of accuracy between j48 decision tree, multilayer perceptron and
naïve bayes has been mentioned.
