16 results about "Food recognition" patented technology

The invention provides a food classification component identification method and system based on multi-scale feature attention synergy, and the method comprises the steps: constructing a multi-task neural network model, sequentially inputting a food image into a global feature module and a progressive local feature module, and respectively extracting global semantic information and refined local features of different scales; feature interaction between food category identification and component analysis tasks is promoted through a cross attention cooperation module, and key areas and important component information in images are dynamically selected; gradually activating different depth levels of the network by adopting a staged end-to-end training strategy, and introducing a KL divergence constraint to enhance the difference between feature scales; and simultaneously generating prediction results of food categories and component labels through a multi-task classifier. The technical problems of food identification and component analysis task splitting, insufficient local feature expression and insufficient cross-task information complementation are solved.

The invention provides an intelligent dinner plate based on multispectral thermal imaging fusion, and belongs to the technical field of intelligent dinner plates. The multi-mode sensing module collects physical and chemical information of food in the dinner plate; the edge calculation and main control module executes real-time data processing at an equipment end; the lightweight food identification and segmentation model is used for identifying food types based on multispectral image data and segmenting areas of the food; the multi-sensor data fusion algorithm is used for fusing the food types, the area proportion information and the total weight, estimating the initial weight of each kind of food based on a pre-trained density factor, and outputting a structured list containing the food types and the initial weights; the cloud nutrition calculation engine receives the structured list and the related hyperspectral feature data, and queries and identifies standard nutritional ingredients of the food from a nutrition database; and calibrating the initial weight by using the hyperspectral characteristic data, and calculating and summarizing the total nutritional ingredients and total calorie of the whole meal based on the calibrated weight and the finely-adjusted nutritional ingredients.

The invention provides a food information processing method and device, and the method comprises the steps: receiving a to-be-recognized food image uploaded by a user and a question message initiated by the user; generating an answer message corresponding to the question message according to the propaganda copywriting in the food image to be identified, and pushing the answer message to the user; and generating an evaluation report of the to-be-identified food, pushing a report viewing control to the user, and pushing the evaluation report to the user in response to the report viewing control triggered by the user. According to the embodiment, the question message of the user can be replied according to the image of the to-be-recognized food, and the complete evaluation report for the to-be-recognized food is generated, so that the food recognition efficiency is improved, and the user experience is improved while the food safety is guaranteed.

The utility model discloses a food processor system for food identification and heat energy intake control, and particularly relates to the field of food, the food processor system comprises a lifting mechanism, the bottom end of the lifting mechanism is fixedly connected with a weighing mechanism, the weighing mechanism comprises a base, the inner wall of the base is connected with a storage box in a sliding mode, and the storage box is connected with the lifting mechanism. The inner wall of the base is fixedly connected with a limiting frame, the upper surface of the limiting frame is fixedly connected with a weighing sensor, the upper end of the weighing sensor is fixedly connected with a placing frame, and the inner wall of the placing frame is provided with a limiting groove; according to the utility model, the heat of food can be accurately identified through the weighing mechanism, the weight of the food can be detected by the weighing sensor during use, the heat of the food can be accurately measured and calculated under the combination of the species and the weight of the food, and soil and water on the food can be collected into the storage box through the filter screen, so that the food can be conveniently stored. Therefore, the weighing accuracy can be ensured, then centralized processing is carried out, and the problem that a traditional detection table is inconvenient to clean is solved.

Cooking appliance (10) with a cooking chamber (12) in which several slots (E1, ... E9) are defined and a camera (30) which serves to detect at least one cooking parameter of food which is placed in one of the slots, characterized by a light guide (32) which extends between the camera (30) and the cooking chamber (12), wherein the light guide (32) has an optical interface (34) for each slot.

The application is suitable for the technical field of pre-prepared food reheating, and provides a pre-prepared food segmented reheating method and system based on real-time food recognition, which acquires food packaging recognition information, acquires food types and cooking requirements according to the recognition information, compares and corrects an optimal reheating curve with performance parameters of a current device, calculates a curve fitting degree that can be reached by the device, maximizes the approximation of the optimal reheating effect within the range allowed by the device for the corrected reheating curve, acquires individualized parameters through user input or historical preference data, converts the individualized parameters into correction coefficients, dynamically adjusts the reheating curve according to the correction coefficients, and determines target parameters of each stage, generates a feedback vector according to the score combined with operation data, updates parameters of the optimal reheating curve of the corresponding food according to the feedback vector, and optimizes a public optimal curve model according to common characteristics, which has the advantages of intelligence, individualization and self-optimization of the pre-prepared food reheating process.

A method for treating food in an appliance (14) by at least one treatment device of the appliance (14) comprises the steps of a. providing a food recognition system with at least one camera image of a food item to be treated by the appliance (14)(S1), b. picture data about the food item by extracting a number of at least one characteristic features of the food item from the at least one camera image (S2), c. assigning a food identification to the food item based on at least one of the number of characteristic features (S4, S6), d. executing a food treatment process or program allocated to the assigned food identification (S8). According to the invention, the camera image is provided during step a. (S1) by an external camera device (12, 14), wherein the camera image provision is executed—by data retrieval by the food recognition system, and / or—by data transfer by the mobile camera device (12, 20), and / or—on the operator's initiation. Further disclosed is a food treatment system comprising an appliance (14) for treating food and an external camera device (12, 20) as well as a software application.

This disclosure provides a method and apparatus for food ingredient identification, an electronic device, and a computer-readable storage medium, relating to the fields of image recognition and health technology. The method includes: receiving an image to be identified; identifying the food ingredients in the image using a pre-trained food recognition model to obtain a preliminary identification result; obtaining the feature vector of the image to be identified; determining the feature similarity between the feature vector of the image to be identified and the feature vectors of historical images, where historical images are images where food ingredients have been identified and the identification results have been corrected; obtaining similar images from the historical images based on the feature similarity; and correcting the preliminary identification result based on the correction data of the similar images to obtain the final identification result of the food ingredients in the image to be identified. This disclosure improves the accuracy of food ingredient identification by correcting the preliminary identification result of the food ingredients in the image using similar images.

This invention discloses a method for food image segmentation and removal based on instance segmentation and spatial prior clustering, comprising three steps: Step 1, inferring from the photo using the YOLO instance segmentation model, extracting the segmentation mask, and cropping it into a black background image; Step 2, extracting the foreground from the image using the GrabCut algorithm, converting it to a transparent background, and generating an RGBA format image; Step 3, performing spatial partitioning on the transparent background image based on Euclidean distance transformation, using the outer perimeter as the plate prior and the center as the food prior, performing K-means clustering in the Lab color space, automatically identifying the plate and food clusters based on the spatial enrichment ratio of each cluster at the perimeter and center, and removing plate pixels based on external reachability connectivity verification, outputting a transparent background image containing only the food. This invention eliminates the need for manual region labeling or setting color thresholds, can adaptively handle plates of different colors, shapes, and sizes, is suitable for diverse dining scenarios, and supports tasks such as food recognition and nutritional estimation.

Food image recognition has been widely applied to scenes such as intelligent ordering, but the problems of insufficient recognition precision and low calculation efficiency of Chinese food due to the characteristics of changeable food material forms, similar appearances of different dishes and the like are still solved. The invention provides a Chinese food lightweight identification method based on improved YOLOv8n. A DCNv3 improved C2f module is introduced into a backbone network, so that the modeling capability of food form change details is enhanced; the CAA-HSFPN feature pyramid is fused in Neck, and the multi-scale feature fusion effect is improved, so that similar dishes are effectively distinguished, and the model parameter quantity is reduced. A DCA-YOLOv8n detection network is constructed on the basis of improvement, and the effectiveness of the detection network is verified through comparison experiments with various mainstream models on a self-established Chinese food data set. According to the method, the performance indexes of Chinese food identification accuracy and the like are remarkably improved while the light weight is kept, and the method is deployed in an actual Chinese food identification system.

A presentation system S includes: a user information recognition unit configured to recognize user information; a current emotion recognition unit configured to recognize a current emotion of a user; a recommended food recognition unit configured to recognize a recommended food item by using a prediction model which receives an input of the user information and the current emotion and outputs the recommended food item; and a food information presentation unit configured to obtain and present the user with food information about the recommended food item. The current emotion recognition unit is configured to estimate, as the current emotion, coordinates selected by the user from an emotion model in which coordinates are defined based on a plurality of basic emotions.

The application discloses a method for intelligent food recognition and nutrition analysis based on deep learning, and belongs to the technical field of image analysis.The method comprises the following steps: 1, collecting input data and pre-processing the data; 2, determining the position of a target in a scene and a mask area, and generating a dense depth map; 3, constructing a food three-dimensional point cloud, denoising the three-dimensional point cloud, and removing abnormal points; 4, obtaining three-dimensional features, extracting two-dimensional features from the image, and performing weight fusion; 5, splicing the obtained features, and strengthening the feature information after fusion; 6, regressing and predicting food weight and heat; and 7, optimizing model prediction and verifying the prediction effect.The method adopts a 2D+3D modal feature fusion mode, performs data enhancement on an original image, extracts 2D features, simultaneously generates a 3D point cloud, and performs multi-modal feature fusion to predict food heat and weight, so that the method can help consumers accurately estimate food heat, weight and other indexes, and formulate scientific diet methods and suggestions.

The invention relates to the technical field of medical health and intelligent diet management, in particular to an intelligent diet intervention system based on real-time food recognition and blood glucose analysis. Comprising a food identification module for collecting food images, identifying food types and extracting image feature information; the nutrition analysis module is used for calculating nutritional ingredients and carbohydrate content based on the food identification result; the blood glucose processing module is used for collecting, calibrating and analyzing continuous blood glucose data and generating blood glucose trend information; the prediction decision-making module is used for predicting the blood glucose change according to the nutrition data and the blood glucose information and generating a personalized intervention scheme; and the data management module executes data storage, encryption protection, information display and user interaction. The complete intelligent diet intervention system is formed by organically combining real-time food recognition, nutrition analysis, blood sugar processing, blood sugar prediction, intervention decision making and data management, and real-time monitoring and dynamic management of postprandial blood sugar changes of the user are achieved.

An object of a system according to an embodiment is to efficiently manage the contents of a refrigerator and reduce food waste while maintaining the balance of meals.SOLUTION: A system according to an embodiment includes an ingredient recognition unit, a recipe suggestion unit, a nutrient analysis unit, an ingredient ordering unit, and a food waste reduction unit. The food recognition unit recognizes the contents of the refrigerator. The recipe proposal unit proposes a recipe suitable for the mood and health goal of the day on the basis of the ingredient recognized by the ingredient recognition unit. The nutrition analysis unit analyzes the nutritional components of the recipe proposed by the recipe proposal unit and checks the balance of the meal. The ingredient ordering unit orders necessary ingredients based on the result checked by the nutrition analysis unit. The food waste reduction unit manages the use-by date of the ingredient recognized by the ingredient recognition unit and makes a proposal for reducing the food waste.SELECTED DRAWING: Figure 1

The invention relates to the technical field of supply chains, in particular to a chain buffet catering supply chain management method and system, and aims to predict catering food sales volume in the future three days in non-holidays through an intelligent sales volume prediction model, reduce catering food waste and excessive supply, and improve catering food supply efficiency. A clustering model is adopted to determine the overlapping range of a densely populated area and areas with high historical sales volume of buffet restaurants in each segmented area block in festivals and holidays, and then the food sales volume of the self-service restaurants with high historical sales volume in the overlapping range is predicted through a festival and holiday intelligent sales volume prediction model, and supply is carried out according to the sales volume. A camera is installed in a storage warehouse of each store, and a catering food recognition model is combined to recognize the freshness degree and consumption degree of supplied catering food, so that the situation that the catering food cannot be supplied due to the occurrence of stale food and emergencies is prevented, the cost is reasonably controlled, and the working efficiency is improved. And subsequent income reduction caused by food safety problems or short supply of catering food is prevented.