CN-122022171-A - Cooking process AI analysis and evaluation method and system

CN122022171ACN 122022171 ACN122022171 ACN 122022171ACN-122022171-A

Abstract

The invention provides an AI analysis and evaluation method and system for a cooking process, which relate to the technical field of analysis and evaluation, wherein a multi-view sensor array is arranged, data acquisition and processing are carried out on the cooking process according to the multi-view sensor array to obtain cooking acquisition processing data, key feature extraction processing is carried out on the cooking acquisition processing data to obtain key feature extraction data, cooking process analysis is carried out on the key feature extraction data to obtain cooking process analysis data, AI analysis operation process judgment is carried out on the cooking process analysis data to obtain operation process judgment information, finished dish information is obtained, and AI analysis and evaluation are carried out on the finished dish information combined with the operation process judgment information to obtain AI analysis and evaluation data. The invention can realize the quantitative analysis and accurate positioning of the core influence links of the operation and the quality association of the finished product.

Inventors

SHU HUAJUN
LIN LINGZHEN

Assignees

广州市锐星信息科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260202

Claims (10)

1. A cooking process AI analysis and evaluation method, characterized in that the method comprises: S1, setting a multi-view sensor array, and acquiring and processing data according to the cooking process by the multi-view sensor array to obtain cooking acquisition and processing data; S2, carrying out key feature extraction processing on the cooking acquisition processing data to obtain key feature extraction data, and carrying out cooking process analysis on the key feature extraction data to obtain cooking process analysis data; s3, carrying out AI analysis operation process judgment on the cooking process analysis data to obtain operation process judgment information; And S4, acquiring finished dish information, and carrying out AI analysis and evaluation on the finished dish information combined with the operation process judgment information to obtain AI analysis and evaluation data.
2. The cooking process AI analysis and evaluation method according to claim 1, wherein S1 comprises: performing region division on a monitoring region of a cooking process according to a functional partition of a cooking scene to obtain a plurality of functional monitoring regions; setting a multi-view sensor array for each function monitoring area, and analyzing and controlling acquisition parameters of the multi-view sensor array according to a preset cooking task to obtain acquisition parameter control data; Synchronously acquiring video stream data in the cooking process through acquisition parameter control data; converting video stream data into a unified coding format, and converting various sensing data into a numerical matrix; And carrying out data fusion processing on the numerical matrix by a data fusion method to generate cooking acquisition processing data.
3. The AI analysis and evaluation method for a cooking process according to claim 2, wherein the step of setting a multi-view sensor array for each function monitoring area, analyzing and controlling acquisition parameters of the multi-view sensor array according to a preset cooking task, and obtaining acquisition parameter control data comprises the steps of: Extracting key information of a preset cooking task to obtain task key extraction information; Collecting parameter analysis is carried out on the task key extraction information, and task collecting analysis data are obtained; generating an acquisition parameter control instruction according to the task acquisition analysis data; and carrying out acquisition parameter control on each area sensor according to the acquisition parameter control instruction to obtain acquisition parameter control data.
4. The cooking process AI analysis and evaluation method according to claim 1, wherein S2 comprises: carrying out key feature recognition of an evaluation target on the cooking acquisition processing data to obtain key feature recognition data; classifying the key feature identification data according to the preset data acquisition types to obtain key feature class data; Performing feature depth extraction on the key feature identification data to obtain feature depth extraction data; performing time sequence association analysis on the feature depth extraction data, constructing a trend curve of the feature changing along with time, and acquiring a feature mutation time node according to the trend curve; matching the characteristic mutation time node with a preset cooking stage division standard to obtain a node stage matching result; determining cooking stage analysis data according to the node stage matching result, and generating cooking process analysis data according to the plurality of cooking stage analysis data.
5. The AI analysis and evaluation method for a cooking process according to claim 4, wherein the performing time-series correlation analysis on the feature depth extraction data to construct a trend curve of feature variation with time, and obtaining a feature mutation time node according to the trend curve, comprises: Acquiring key feature change data according to the feature depth extraction data; Constructing a trend curve of the characteristic change along with time according to the key characteristic change data; comparing the key feature change data with a feature change threshold value to obtain a feature change comparison result; Determining characteristic change nodes of the trend curve according to the characteristic change comparison result; The method comprises the steps of obtaining the duty ratio of the total number of feature change nodes in a feature sum, and obtaining a feature change duty ratio coefficient; Comparing the characteristic change duty ratio coefficient with a preset characteristic change duty ratio threshold value to obtain a change duty ratio comparison result; And determining the characteristic mutation time node according to the comparison result of the change duty ratio.
6. The cooking process AI analysis and evaluation method according to claim 1, wherein S3 includes: constructing an operation flow judgment model through a fusion architecture of a convolutional neural network and a long-term and short-term memory network; Inputting the cooking process analysis data into an operation flow judgment model, performing operation characteristic flow matching on a plurality of cooking stages through the operation flow judgment model, and outputting a matching result; Calculating an operation compliance score according to the matching result to obtain a stage operation coefficient; Performing stage operation abnormal positioning according to the stage operation coefficients to obtain stage abnormal positioning data; and carrying out operation process judgment according to the stage abnormal positioning data to obtain operation process judgment information.
7. The cooking process AI analysis and evaluation method according to claim 6, wherein the performing operation process determination based on the stage abnormality positioning data to obtain operation process determination information includes: Acquiring an operation process judgment index and determining an index weight of the operation process judgment index; Performing index anomaly analysis on the stage anomaly positioning data according to the operation process judgment index and the index weight to obtain index anomaly analysis data; Comparing the index anomaly analysis data with a preset index anomaly threshold value to obtain an index anomaly comparison result; And carrying out operation process judgment according to the index abnormality comparison result to obtain operation process judgment data.
8. The cooking process AI analysis and evaluation method according to claim 1, wherein S4 includes: Acquiring multi-angle images of the finished dish through a multi-angle sensor array to obtain the information of the finished dish; extracting features of the finished product vegetable information to obtain finished product feature data of the finished product vegetable information; And constructing an AI comprehensive evaluation model, inputting the characteristic parameters and the operation process judgment information of the finished dish into the AI comprehensive evaluation model, and generating AI analysis evaluation data through the AI comprehensive evaluation model. Preprocessing the information of the finished dish, and extracting quality characteristic data of the finished dish; Establishing a mapping relation between an operation process and quality characteristic data according to the operation process judging information to obtain operation influence relation information; and acquiring AI analysis and evaluation data according to the operation influence relation information.
9. The cooking process AI analysis evaluation method according to claim 8, wherein the acquiring AI analysis evaluation data based on the operation influence relation information includes: analyzing the operation influence relation information to obtain a plurality of corresponding association strengths and generating an association strength matrix; setting weight coefficients for the quality characteristic data according to preset cooking quality evaluation standards; combining the correlation intensity matrix and the weight coefficient, calculating the influence weight of each operation link on the whole quality of the finished dish, and obtaining influence weight data; The extraction operation process judging information is combined with influence weight data to generate a negative influence quantification report on the quality characteristic data; and inputting the negative influence quantification report and the quality characteristic data into a constructed AI comprehensive evaluation model, and outputting AI analysis evaluation data.
10. A cooking process AI analysis and evaluation system, the system comprising: The data monitoring module is used for setting a multi-view sensor array, and carrying out data acquisition and processing according to the cooking process of the multi-view sensor array to obtain cooking acquisition processing data; the process analysis module is used for carrying out key feature extraction processing on the cooking acquisition processing data to obtain key feature extraction data, and carrying out cooking process analysis on the key feature extraction data to obtain cooking process analysis data; the operation judging module is used for carrying out AI analysis operation process judgment on the cooking process analysis data to obtain operation process judgment information; and the finished product evaluation module is used for acquiring finished product dish information, and carrying out AI analysis and evaluation on the finished product dish information combined with the operation process judgment information to acquire AI analysis and evaluation data.

Description

Cooking process AI analysis and evaluation method and system Technical Field The invention provides a cooking process AI analysis and evaluation method and system, relates to the technical field of analysis and evaluation, and particularly relates to the technical field of cooking process AI analysis and evaluation. Background Under the background of increasingly growing requirements for standardized management and control and cooking skill improvement in the catering industry, the cooking process and the evaluation of finished dishes become key links. The traditional cooking evaluation mode depends on manual subjective judgment, so that the efficiency is low, the labor cost is high, and evaluation deviation caused by individual difference exists. Meanwhile, the prior art is difficult to establish accurate association between the cooking operation process and the quality of the finished product dish, the influence degree of different operation links on the quality of the finished product cannot be quantified, and the core operation links causing the quality problem of the finished product are difficult to accurately position. The cooking optimization lacks a clear direction, the operation standard management and control is difficult to fall to the ground, and the stability improvement of the quality of dishes and the efficient inheritance of cooking skills are restricted. Disclosure of Invention The invention provides an AI analysis and evaluation method and system for a cooking process, which are used for solving the problems: The invention provides a cooking process AI analysis and evaluation method and a cooking process AI analysis and evaluation system, wherein the method comprises the following steps: S1, setting a multi-view sensor array, and acquiring and processing data according to the cooking process by the multi-view sensor array to obtain cooking acquisition and processing data; S2, carrying out key feature extraction processing on the cooking acquisition processing data to obtain key feature extraction data, and carrying out cooking process analysis on the key feature extraction data to obtain cooking process analysis data; s3, carrying out AI analysis operation process judgment on the cooking process analysis data to obtain operation process judgment information; And S4, acquiring finished dish information, and carrying out AI analysis and evaluation on the finished dish information combined with the operation process judgment information to obtain AI analysis and evaluation data. Further, the step S1 includes: performing region division on a monitoring region of a cooking process according to a functional partition of a cooking scene to obtain a plurality of functional monitoring regions; setting a multi-view sensor array for each function monitoring area, and analyzing and controlling acquisition parameters of the multi-view sensor array according to a preset cooking task to obtain acquisition parameter control data; Synchronously acquiring video stream data in the cooking process through acquisition parameter control data; converting video stream data into a unified coding format, and converting various sensing data into a numerical matrix; And carrying out data fusion processing on the numerical matrix by a data fusion method to generate cooking acquisition processing data. Further, the setting of the multi-view sensor array for each function monitoring area, analyzing and controlling the acquisition parameters of the multi-view sensor array according to the preset cooking task, and obtaining the acquisition parameter control data includes: Extracting key information of a preset cooking task to obtain task key extraction information; Collecting parameter analysis is carried out on the task key extraction information, and task collecting analysis data are obtained; generating an acquisition parameter control instruction according to the task acquisition analysis data; and carrying out acquisition parameter control on each area sensor according to the acquisition parameter control instruction to obtain acquisition parameter control data. Further, the step S2 includes: carrying out key feature recognition of an evaluation target on the cooking acquisition processing data to obtain key feature recognition data; classifying the key feature identification data according to the preset data acquisition types to obtain key feature class data; Performing feature depth extraction on the key feature identification data to obtain feature depth extraction data; performing time sequence association analysis on the feature depth extraction data, constructing a trend curve of the feature changing along with time, and acquiring a feature mutation time node according to the trend curve; matching the characteristic mutation time node with a preset cooking stage division standard to obtain a node stage matching result; determining cooking stage analysis data according to the node stage matching result, and gen