Tomato Grading: A New Approach for Classifying and Predicting Tomato Quality based on Visual Features

Abstract

Increased awareness about nourishing and healthy lifestyles to propel the consumption of vegetables in order to meet diverse dietary and nutritional needs. The global tomato market was valued to register a Compound Annual Growth Rate of more than 3.8% over the projection horizon of 2021-2026. The planned approach that calculates the grade of the tomato in regard to its external features. Grading is sorting or categorization of tomatoes into different grades according to the size, shape, colour etc and is one of the foremost necessary processes in post harvesting, however this procedure is sometimes administered manually, that is not economical as a result it needs huge estimate of enrollment, and have an inclination to human error. The grading method is performed by capturing the tomato image using web camera which calculates the percentage of ripeness based on unique set of features that are utilized to train the neural network. Color emerges as an extremely prominent feature for recognizing defect and matureness of the tomato. The major objective is to check the tomato quality with high speed for evaluating maximal count of tomatoes in least amount of time. For spoiled tomatoes, the proposed system helps in identification of tomato plant disease and allocate countermeasures that can be used as a fortification mechanism against the disease. The tomato plant disease detection can be done by observing the spots on the leaves of the diseased plant. In order to detect plant diseases, the approach we are endorsing is image processing using Convolution neural network (CNN).