Red meat classification method and device based on deep learning, equipment and storage medium

Abstract

The invention discloses a red meat classification method and device based on deep learning, equipment and a storage medium. The method comprises the steps of obtaining to-be-classified red meat hyperspectral image information; inputting the to-be-classified red meat hyperspectral image information into a preset metric function to obtain similarity information among feature information in the to-be-classified red meat hyperspectral image information; extracting reference feature information in the to-be-classified red meat hyperspectral image information according to the similarity information;performing dimensionality reduction on the reference feature information to obtain target feature information; and performing spatial-spectral joint classification on the target feature information based on deep learning to obtain target type information, thereby firstly extracting feature information of the red meat hyperspectral image information through an appropriate preset metric function, and then extracting the spatial-spectral joint feature information based on deep learning to improve the classification precision of the red meat hyperspectral image.

Description

technical field [0001] The present invention relates to the field of remote sensing technology, in particular to a red meat classification method, device, equipment and storage medium based on deep learning. Background technique [0002] Hyperspectral technology developed in the field of remote sensing in the past decade has been applied to non-destructive testing in many fields. Compared with other images, hyperspectral images contain rich spectral information in addition to spatial information. Combining spectral processing methods and image processing The algorithm is suitable for non-destructive testing of red meat quality. At present, some achievements have been made in the detection of beef, mutton and pork based on hyperspectral technology, including tenderness, pH value, water strength, marbling and microbial chemistry; [0003] However, hyperspectral images have the characteristics of high dimensionality, spectral similarity and mixed pixels, and the amount of data...

AI Technical Summary

Problems solved by technology

[0004] The main purpose of the present invention is to propose a red meat classification method, device, equipment and storage medium based on deep learning, aiming to solve the technical problem of low recognition accuracy when classifying hyperspectral images

Images