Author : Amruta Supekar and Prof. Madhuri Wakode

Mangoes are one of the most favorite fruits in the world. A large scale of mangoes is globally exported and also locally consumed. During mango exports and local marketing, quality inspection is essential. It directly affects customer satisfaction and thereby vendor’s monetary gains. Mango grading is a postharvest operation in which mangoes are classified according to grading parameters like size, shape, ripeness, defect, sweetness, etc. This process of quality assessment is usually done manually just by inspection with a naked eye. However, it can lead to inconsistent and inaccurate grading. A computer vision-based automated system, utilizing a mango image can provide a better, accurate, consistent and reliable solution. In this research work, an attempt to grade Kesar mangoes based on its maturity, size, and shape has been made.  Firstly, the mango image is analyzed to determine its size, ripeness, and shape category. Further, three parameters are combined using pre-defined grading rules to determine mango grade as rejected, grade1 or grade2 quality.

Keywords: Computer Vision; Image Processing; Machine Learning;  Mango Grading; Mango Classification