11 results about "Food classification" patented technology

The application provides a model training method, a package image recognition method and a device thereof. The model training method comprises: obtaining label information corresponding to a first package image, the first package image being a package image of a first sample food; analyzing the label information to obtain a first text; replacing at least part of the first text with a target preset text to obtain a second text, the target preset text having an association with the first text, and the second text being a description of a second sample food; and training a food classification model according to the first text and the second text to obtain a trained food classification model. The embodiment of the application replaces at least part of the first text on the first package image with the target preset text, expands the training sample, avoids the need for manual labeling of characters on package images of various foods to obtain sufficient training samples, thereby reducing the labeling workload of the staff and improving the efficiency of model training.

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 discloses a food classification deep learning method adaptive to a few-sample scene, and the method comprises the steps: obtaining a few-sample food classification data set, carrying out the preprocessing enhancement of an original image in the few-sample food classification data set, obtaining an image x, constructing an RL-Clip model, carrying out the feature processing of the image x, and finally obtaining a food classification probability matrix. Secondly, optimizing an image processing process; the purposes of improving the generalization of the detection model and improving the detection effectiveness, accuracy and confidence of the specific data set are achieved.

The objective is to provide a diversity assessment device, a diversity assessment method, and a diversity assessment program that can assess food diversity while considering nutrient diversity by evaluating the diversity of ingested foods based on clusters that classify foods. [Solution] Based on food component data, a food classification is set by classifying the target specified foods according to nutrients using a clustering method, data on the intake status of specified foods included in a meal classified according to the food classification is obtained, and a dietary diversity score is calculated for the specified food intake status data.

The utility model discloses a pre-packaged food inventory classification management frame, and relates to the technical field of food classification management. The device comprises a plurality of layering plates, wherein a splicing mechanism and a classifying mechanism are arranged on the layering plates; the splicing mechanism comprises an adjusting assembly, a supporting assembly and a splicing assembly, the adjusting assembly comprises a supporting column arranged on the layering plate, the layering plate is fixedly connected with a plurality of sliding blocks, the sliding blocks are provided with spring grooves, the spring grooves are fixedly connected with springs, and the springs are fixedly connected with bearing blocks. By means of the splicing mechanism and the elasticity of the spring, a worker can assemble and limit the layering plate and the supporting columns by pressing the ejector block, meanwhile, when the layering management height of the layering plate needs to be adjusted, rapid adjustment can be completed by pressing the ejector block, two management frames are spliced through cooperation of a clamping ring and a clamping block, and the labor intensity of workers is lowered. And the device is suitable for storing various pre-packaged foods with different sizes, the application range of the device is widened, and the convenience of the device is improved.

The utility model relates to the technical field of storage tanks, and discloses a food-grade multifunctional storage tank which comprises a storage tank body, a clamping groove is formed in the inner wall of the storage tank body, a flow dividing assembly is installed in the storage tank body, a sealing cover assembly is connected to the top of the storage tank body in a clamped mode, and the flow dividing assembly comprises a baffle. A sliding block is fixedly connected to the surface of the baffle, a flow dividing barrel is slidably connected to the surface of the sliding block, a sliding groove is formed in the surface of the flow dividing barrel, a lifting handle is fixedly connected to the end, away from the sliding groove, of the flow dividing barrel, a threaded sleeve is fixedly connected to the top of the baffle, and a bolt is in threaded connection with the interior of the threaded sleeve. According to the food storage container, by arranging a matching structure of the flow dividing cylinder, the sliding block and the sliding groove, rapid extraction and resetting of the flow dividing cylinder are achieved, in combination with the detachable design of the sealing plate, it is guaranteed that different kinds of food are stored in a classified mode, independent taking and using are convenient, and the problem that food is mixed or tainted by other odors in a traditional storage container is solved.

This application relates to the field of smart home appliance technology, and discloses a method and apparatus for cross-domain small-sample object image classification for smart terminals. The method includes: acquiring an image of food to be identified; processing the image to extract basic feature vectors; mapping the basic feature vectors using a dual Riemannian manifold processing module to obtain Euclidean space features and hyperbolic space features respectively; fusing the Euclidean space features and hyperbolic space features to obtain a comprehensive feature representation for food classification; and determining the food classification result based on the comprehensive feature representation. The dual Riemannian manifold processing module includes Euclidean space branches and hyperbolic space branches. This method achieves multi-space feature fusion by collaboratively extracting Euclidean space features and hyperbolic space features using dual Riemannian manifolds, thereby improving the feature representation capability and classification accuracy of food images.

The utility model discloses intelligent cold-chain food grading equipment with a rapid classification function, and relates to the technical field of cold-chain food classification, the technical scheme is as follows: the intelligent cold-chain food grading equipment comprises a support frame and a plurality of buffer components arranged on the support frame, each buffer component comprises a support plate, the support plate is mounted on one side of the support frame in a mirroring manner, and the support plate is arranged on the other side of the support frame; each two adjacent supporting plates form a group, a conveying plate is slidably mounted between each group of supporting plates, a plurality of sliding grooves are formed in each conveying plate, and supporting rods are symmetrically mounted below each group of supporting plates. According to the cold-chain food packaging machine, the problems that when food falls off, large impact force is generated, the impact force can directly act on food packages, if the impact force is too large, the cold-chain food is broken or deformed, and meanwhile the higher the falling height is, the larger the impact force is, the more vertical the falling angle is, and the larger the impact force is are solved.

The disposable meal box comprises a box bottom, the upper side of the box bottom is connected with two first side face parts and two second side face parts in a turnover mode, and two triangular face parts are connected between the adjacent first side face parts and the second side face parts in a foldable mode. The upper side of the second side face part is provided with a buckling and folding face part used for fixing the whole in a foldable mode, two partition plates are arranged on the upper side of the box bottom, and through the arrangement of the buckling and folding face part and the triangular face part, the first side face part and the second side face part can form a stable storage space. The box cover part is matched to form a sealing effect and a stable lunch box whole body; when the meal box needs to be folded and stored, the whole meal box can be folded into a flat structure only by taking down the clamping pieces, the turnover partition plate, the first side face part and the second side face part, the storage size of the meal box after eating is greatly reduced, and leakage of remaining meal is avoided.

This application provides a deep learning-based food classification method, system, and readable storage medium. The method includes acquiring a training dataset, which comprises a first time-series spectral signal corresponding to a pure sample type and a second time-series spectral signal corresponding to an adulterated sample type; using the training dataset as the original time series and transforming the original time series to obtain a first-order difference sequence and a Fourier sequence; performing two-dimensional processing on each sequence using a tiling and sliding window approach to obtain corresponding two-dimensional data; constructing an initial classification model and inputting the obtained two-dimensional data and Fourier sequences into the initial classification model for training, obtaining a target classification model upon training termination; and inputting the acquired two-dimensional data to be processed and the Fourier sequences into the target classification model to obtain the food classification result. The implementation of this method can improve the accuracy of food classification.