Grapefruit granulation classification method based on convolutional neural network and hyperspectral technology

Abstract

The invention relates to a grapefruit granulation classification method based on a convolutional neural network and a hyperspectral technology. The grapefruit granulation classification method comprises the following steps of 1) obtaining grapefruit diffuse transmission hyperspectral information, 2) performing normalization preprocessing on the obtained hyperspectral information, and then dividingthe processed information into a training set and a verification set, 3) training the established multi-layer convolutional neural network model by using the training set, 4) establishing a loss function, training the multi-layer convolutional neural network model by adopting an Adagrad gradient descent mode in combination with back propagation, and taking the model with the minimum loss as the trained multi-layer convolutional neural network model, and 5) inputting the training set and the verification set into the trained convolutional neural network model to obtain a classification result.The method is beneficial to nondestructive detection of grapefruit granulation degree and classification of grapefruit granulation degree.

Description

technical field [0001] The invention belongs to the technical field of fruit granulation detection, in particular to a grapefruit granulation classification method based on convolutional neural network and hyperspectral technology. Background technique [0002] Citrus is an evergreen subtropical fruit tree and is the most important economic tree fruit in the world. It is grown commercially in many countries. Juice granulation is a common physiological phenomenon of citrus fruit trees. It is a physiological disease that occurs during the late ripening period and postharvest storage. Its main manifestations are abnormal swelling and hardening of juice sacs, lignification of juice sacs, and loss of juice flavor. , resulting in a decline in the edible quality of the fruit, or even loss of commodity value, bringing great economic losses to the fruit growers. However, the degree of granulation of grapefruit cannot be observed by external light, and it is usually judged by breakin...

AI Technical Summary

Problems solved by technology

Juice granulation is a common physiological phenomenon of citrus fruit trees. It is a physiological disease that occurs during the late ripening period and postharvest storage. Its main manifestations are abnormal swelling and hardening of juice sacs, lignification of juice sacs, and loss of juice flavor. , resulting in a decline in the edible quality of the fruit, or even loss of commodity value, bringing great economic losses to fruit farmers

However, the degree of granulation of grapefruit cannot be observed by external light, and it is usually necessary to break the skin of grapefruit to judge

Traditional granulation detection methods are all destructive detection, and non-destructive granulation detection cannot be realized

Images