Density estimation-based kitchen dirt and disorder evaluation method

Abstract

The invention provides a density estimation-based kitchen dirt and disorder evaluation method. The method comprises the following steps: S1, collecting a kitchen video, and extracting a kitchen image X from a video frame corresponding to the kitchen video; s2, manually marking a dirty and disorder region in the kitchen image X to form a marking frame with the size within a preset range, and scoring the dirty and disorder condition; s3, constructing a dirty and disorder density map M with the same size as the kitchen image X according to the distribution of the labeling boxes and the labels; performing size processing and normalization on the plurality of dirty and disorder density maps M; s4, constructing and training a neural network for a dirty and disorder evaluation algorithm according to image data of the kitchen image X and the dirty and disorder density map M; and s5, evaluating the dirty and disorder degree of the kitchen in the video image by using a neural network. According to the method, the dirty and disorder conditions of the scene are automatically scored through the convolutional neural network, and meanwhile, the dirty and disorder density map is adopted to identify the specific positions of the dirty, messy and bad conditions, so that merchants are helped to improve the kitchen environment.

Description

technical field [0001] The invention relates to the technical field of image analysis, in particular to a density estimation-based method for evaluating kitchen mess. Background technique [0002] In the 2018 notice issued by the State Administration of Market Regulation on the issuance of guidance on the work of bright kitchens and bright stoves in catering services (Guoshi Jianshijian No. 2 [2018] No. 32), in order to supervise catering service providers, strengthen food safety management, and standardize the processing process, To promote social co-governance of food safety in catering services, according to the relevant provisions of the "Food Safety Law of the People's Republic of China", the State Administration for Market Regulation has formulated the "Guiding Opinions on the Work of Bright Kitchens and Stoves in Catering Services". As an important item in the bright kitchen, the evaluation and evaluation of the mess in the back kitchen is the current focus of the Mar...

AI Technical Summary

Problems solved by technology

[0004] 1. Floor stains

[0005] 2. Oil stains (stains) on the wall

[0006] 3. Damaged walls and floors

[0007] 4. Ground debris food residue

[0008] 5. Complex debris on the ground

[0009] 6. Desktop chaos

There is also a data-based heat map generation method, but compared to the messy kitchen, there is not so much data for image drawing

Images