Vegetable quality detection method, system, medium and equipment

Abstract

The invention discloses a vegetable quality detection system which comprises a reading module, a feature marking module, a training learning detection module, a grading module and a result output module. The diagnostic system is stored on a hard disk of a computer. The reading module is used for storing a vegetable quality detection system on a hard disk of a computer and inputting vegetable images on the computer through view transmission equipment; the feature marking module is used for marking each feature of the vegetable image to form a sample; the training learning detection module is used for constructing and training a YOLO deep neural network and detecting vegetable images; and the grading module is used for acquiring the length proportion and the area proportion according to the coordinate information of each prediction frame of the vegetables, and finally comparing the length proportion and the area proportion with a set standard to obtain a grading result. The invention further provides a corresponding medium and equipment, the space of a storage hard disk required by the system is reduced, and the deep learning network model used by the system greatly improves the training speed and the grading efficiency.

Description

technical field [0001] The invention relates to the technical field of target detection, in particular to a vegetable quality detection method, system, medium and equipment. Background technique [0002] Among the existing vegetable identification methods, traditional target detection methods are mostly used to identify vegetables. However, the traditional target detection method has the problem of being time-consuming and cannot meet the real-time requirements of vegetable production. With the continuous development of deep learning technology, people use the deep convolutional network model to detect vegetables, but the current deep convolutional network technology for vegetables detection accuracy is not high, and the real-time detection level is not strong. SUMMARY OF THE INVENTION [0003] In order to overcome the defects and deficiencies of the prior art, the present invention proposes a vegetable quality detection method, which adopts a deep learning network to det...

AI Technical Summary

Problems solved by technology

However, the traditional target detection method has the problem of time-consuming and cannot meet the real-time requirements of vegetable production.

With the continuous development of deep learning technology, people use deep convolutional network models to detect vegetables, but the detection accuracy of current deep convolutional network technology for vegetables is not high, and the real-time detection level is not strong

Images