CN-121997272-A - Agricultural product quality grade grading method, system and electronic equipment

CN121997272ACN 121997272 ACN121997272 ACN 121997272ACN-121997272-A

Abstract

The disclosure relates to the technical field of agricultural product intellectualization, in particular to an agricultural product quality grade grading method, an agricultural product quality grade grading system and electronic equipment. The method comprises the steps of determining appearance grade data of the agricultural products according to appearance image data of the agricultural products, obtaining quality fingerprint data of the agricultural products, wherein the quality fingerprint data represents physicochemical indexes of the agricultural products, the physicochemical indexes comprise at least one of sugar degree, acidity, pesticide residue degree and production place characteristics, and determining quality grade information of the agricultural products according to the appearance grade data and the quality fingerprint data, and the quality grade information represents the quality grade of the agricultural products. According to the technical scheme, accuracy of quality grading of agricultural products is improved.

Inventors

WANG ZHIYONG
LI GUORAN
ZHAO YAN

Assignees

北京新盟风控科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260206

Claims (10)

1. A method for grading quality grades of agricultural products, comprising: determining appearance grade data of the agricultural products according to the appearance image data of the agricultural products; acquiring quality fingerprint data of the agricultural products, wherein the quality fingerprint data represents physicochemical indexes of the agricultural products, and the physicochemical indexes comprise at least one of sugar degree, acidity, pesticide residue degree and production place characteristics; And determining quality grade information of the agricultural products according to the appearance grade data and the quality fingerprint data, wherein the quality grade information represents the quality grade of the agricultural products.
2. The method of claim 1, wherein the obtaining quality fingerprint data of the agricultural product comprises: determining the quality fingerprint data according to the identification information, wherein the quality fingerprint data corresponds to the identification information; or identifying the type information of the agricultural products according to the appearance image data, wherein the type information comprises the type information and the type information of the agricultural products, and inquiring from a preset quality database according to the type information to obtain the quality fingerprint data, wherein the preset quality database stores different types of agricultural products and the quality fingerprint data corresponding to the different types of agricultural products.
3. The method of claim 1, wherein said determining quality grade information for said agricultural product based on said appearance grade data and said quality fingerprint data comprises: Determining the appearance weight and the quality fingerprint weight of the agricultural product according to the category information of the agricultural product, wherein the category information comprises the category information and the category information of the agricultural product; determining a first quality score according to the appearance grade data and the appearance weight; determining a second mass fraction according to the quality fingerprint data and the quality fingerprint weight; and determining quality grade information of the agricultural product according to the first quality fraction and the second quality fraction.
4. A method according to claim 3, wherein said determining the appearance weight of said agricultural product based on the category information of said agricultural product comprises: Determining the initial weight of the appearance of the agricultural product according to the type information of the agricultural product; Determining a confidence coefficient fluctuation of the agricultural product according to the variety information of the agricultural product; And determining the appearance weight of the agricultural product according to the appearance initial weight and the confidence coefficient fluctuation coefficient.
5. A method according to claim 3, wherein said determining a quality fingerprint weight of said agricultural product based on said agricultural product's category information comprises: Determining the quality initial weight of the agricultural product according to the type information of the agricultural product; Determining a timeliness coefficient of the quality fingerprint data according to time information of the quality fingerprint data, wherein the time information is used for representing data storage time of the quality fingerprint data and origin characteristic change information of the agricultural products; and determining the quality fingerprint weight according to the quality initial weight and the timeliness coefficient.
6. The method according to any one of claims 1-5, applied to a server, the method further comprising: Receiving the appearance image data sent by a client, wherein the appearance image data is acquired by the client according to the type information of the agricultural products through an image acquisition module and under the guidance of an image acquisition guide instruction, the image acquisition guide instruction comprises an image frame and illumination prompt information, the image frame is matched with the appearance shape of the agricultural products, and the image acquisition guide instruction is generated by the client according to the type information of the agricultural products; and sending the quality grade information of the agricultural products to the client so that the client generates a quality grading result of the agricultural products according to the quality grade information.
7. A method for grading quality grades of agricultural products, characterized by being applied to a client, the method comprising: According to the type information of the agricultural products, an image acquisition module is started and an image acquisition guide instruction is generated to acquire the appearance image data of the agricultural products, the image acquisition guide instruction is used for guiding the acquisition of the appearance image data of the agricultural products, the image acquisition guide instruction comprises an image frame and illumination prompt information, the image frame is used for representing an image area, and the image frame is matched with the appearance shape of the agricultural products; judging whether the appearance image data is qualified or not by adopting a preset image evaluation standard; if the appearance image data does not accord with the preset image evaluation standard, restarting the image acquisition module and generating an image acquisition guiding instruction so as to acquire the appearance image data of the agricultural product again; if the appearance image data meets the preset image evaluation standard, sending the appearance image data to a server so that the server executes the method of any one of claims 1-6; And receiving the quality grade information of the agricultural products returned by the server, and generating a quality grading result of the agricultural products according to the quality grade information.
8. The method of claim 7, wherein prior to said activating an image acquisition module and generating an image acquisition guidance instruction based on the type information of the agricultural product, the method further comprises: The method comprises the steps of responding to an identification recognition instruction, starting an image acquisition module to acquire a physical identification image of the agricultural product, acquiring identification information of the agricultural product according to the physical identification image, wherein the identification information comprises the type information of the agricultural product; or acquiring the type information of the agricultural products according to the type keywords of the agricultural products input by the user.
9. A quality grade grading system for agricultural products, comprising: the system comprises a client, an image acquisition module, an image acquisition guide instruction, a server and an image processing module, wherein the client is used for starting the image acquisition module according to the type information of the agricultural products and generating the image acquisition guide instruction so as to acquire the appearance image data of the agricultural products, the image acquisition guide instruction is used for guiding the acquisition of the appearance image data of the agricultural products, the image acquisition guide instruction comprises an image frame for representing an image area and illumination prompt information, the image frame is matched with the appearance shape of the agricultural products; The server is used for determining appearance grade data of the agricultural products according to the appearance image data of the agricultural products, acquiring quality fingerprint data of the agricultural products, wherein the quality fingerprint data represents physicochemical indexes of the agricultural products, and the physicochemical indexes comprise at least one of sugar degree, acidity, pesticide residue degree and production place characteristics; The client is further used for receiving the quality grade information of the agricultural products returned by the server and generating quality grading results of the agricultural products according to the quality grade information.
10. An electronic device, comprising: At least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

Description

Agricultural product quality grade grading method, system and electronic equipment Technical Field The disclosure relates to the technical field of agricultural product intellectualization, in particular to a method, a system and electronic equipment for grading agricultural product quality grades. Background In the prior art, the quality classification of agricultural products is performed by analyzing the surface characteristics of the agricultural products mainly depending on deep learning algorithms such as Convolutional Neural Network (CNN) and target detection (YOLO). In this way, erroneous judgment is likely to occur in quality classification of agricultural products, and quality classification cannot be effectively performed for some agricultural products having poor appearance. Disclosure of Invention The embodiment of the disclosure provides a farm product quality grade grading method, a farm product quality grade grading system and electronic equipment, which are used for solving or relieving one or more technical problems in the prior art. As a first aspect of the embodiments of the present disclosure, the embodiments of the present disclosure provide a method for grading quality grades of agricultural products, including: Determining appearance grade data of the agricultural products according to the appearance image data of the agricultural products; Acquiring quality fingerprint data of the agricultural products, wherein the quality fingerprint data represents physicochemical indexes of the agricultural products, and the physicochemical indexes comprise at least one of sugar degree, acidity, pesticide residue degree and production place characteristics; and determining quality grade information of the agricultural products according to the appearance grade data and the quality fingerprint data, wherein the quality grade information represents the quality grade of the agricultural products. In some embodiments, obtaining quality fingerprint data for an agricultural product includes: Acquiring identification information of agricultural products; And determining quality fingerprint data according to the identification information, wherein the quality fingerprint data corresponds to the identification information. In some embodiments, obtaining quality fingerprint data for an agricultural product includes: identifying category information of the agricultural products according to the appearance image data, wherein the category information comprises category information and variety information of the agricultural products; And according to the quality information, inquiring from a preset quality database to obtain quality fingerprint data, wherein the preset quality database stores different quality agricultural products and corresponding quality fingerprint data. In some embodiments, determining quality grade information for the agricultural product based on the appearance grade data and the quality fingerprint data includes: Determining the appearance weight and the quality fingerprint weight of the agricultural products according to the category information of the agricultural products, wherein the category information comprises the category information and the category information of the agricultural products; Determining a first quality score according to the appearance grade data and the appearance weight; determining a second mass fraction according to the quality fingerprint data and the quality fingerprint weight; And determining quality grade information of the agricultural product according to the first quality fraction and the second quality fraction. In some embodiments, determining the appearance weight of the agricultural product based on the item information of the agricultural product includes: determining the initial weight of the appearance of the agricultural product according to the type information of the agricultural product; determining a confidence coefficient of fluctuation of the agricultural product according to the variety information of the agricultural product; and determining the appearance weight of the agricultural product according to the appearance initial weight and the confidence coefficient fluctuation coefficient. In some embodiments, determining the quality fingerprint weight of the agricultural product based on the category information of the agricultural product comprises: determining the quality initial weight of the agricultural product according to the type information of the agricultural product; Determining a timeliness coefficient of the quality fingerprint data according to time information of the quality fingerprint data, wherein the time information is used for representing data storage time of the quality fingerprint data and origin characteristic change information of agricultural products; and determining the quality fingerprint weight according to the quality initial weight and the timeliness coefficient. In some embodiments, applied to the s