CN113177525A - AI electronic scale system and weighing method - Google Patents
AI electronic scale system and weighing method Download PDFInfo
- Publication number
- CN113177525A CN113177525A CN202110583656.0A CN202110583656A CN113177525A CN 113177525 A CN113177525 A CN 113177525A CN 202110583656 A CN202110583656 A CN 202110583656A CN 113177525 A CN113177525 A CN 113177525A
- Authority
- CN
- China
- Prior art keywords
- image
- weighing
- index
- model
- image recognition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005303 weighing Methods 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 43
- 239000013598 vector Substances 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims description 22
- 238000003860 storage Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000009467 reduction Effects 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000004590 computer program Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 235000021022 fresh fruits Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 235000013311 vegetables Nutrition 0.000 description 2
- 235000010724 Wisteria floribunda Nutrition 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 235000012055 fruits and vegetables Nutrition 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000007115 recruitment Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/51—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Library & Information Science (AREA)
- Multimedia (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an AI electronic scale system and a weighing method, wherein the AI weighing method comprises the following steps: pre-establishing an image identification model and an image index of a weighing object; establishing an image score model according to the image identification model and the image index; acquiring a feature vector of a weighing object image, respectively inputting the feature vector into an image recognition model and an image index, and calculating a total similarity score of a target image according to the image recognition model and the image index; and outputting sorting and weighing results according to the calculated similarity score. The system and the weighing method adopt artificial intelligence to carry out image recognition of the weighing object, and establish image indexes based on the image recognition result, thereby effectively improving the weighing efficiency.
Description
Technical Field
The invention relates to the field of weighing, in particular to a weighing method of an AI electronic scale system.
Background
In the fresh fruit and vegetable industry, a weighing scale is a device which is necessary for operation and needs to be used, at present, a bar code scale is mainly used by a merchant, the management of commodities is carried out through PLU codes (Price Look Up code), and each PLU code corresponds to one commodity. When the quantity of store SKUs is more (the commodity type of a typical fresh fruit and vegetable merchant is more than 200), the PLU code is difficult to remember clearly, temporary inquiry is needed when weighing, the weighing time is long, and in order to avoid the phenomenon of queuing in peak periods, weighing platforms and weighing personnel are needed to be added in stores. The above-described weighing scales require the addition of an additional weighing platform and a weigher, thereby increasing the cost of the operator. In addition, the traditional PLU codes are complex, a weigher needs to spend time learning and memorizing, and the weigher is easy to forget and make mistakes, so that the labor cost of the operator is increased.
Disclosure of Invention
One of the objects of the present invention is to provide an AI electronic scale system and a weighing method, which perform image recognition of a weighed object by using artificial intelligence, and establish an image index based on the result of the image recognition, thereby effectively improving the weighing efficiency.
One of the purposes of the invention is to provide an AI electronic scale system and a weighing method, the system and the method establish similarity sorting of weighing objects through image recognition, output the front row of the weighing objects with high similarity, a weigher can directly select from the commodities output in the front row without retrieving the weighed commodities, and the weighing efficiency can be effectively improved.
One of the objects of the present invention is to provide an AI electronic scale system and a weighing method, which do not require a weigher to learn, and the weigher can handle a plurality of AI electronic scales at the same time, thereby effectively reducing operator recruitment and training costs.
One of the objectives of the present invention is to provide an AI electronic scale system and a weighing method, where the system and the method can perform mechanical learning in a cloud, and the cloud transmits a learning resolution result to an AI electronic scale, so as to reduce hardware cost.
One of the objects of the present invention is to provide an AI electronic scale system and a weighing method, which can solve the above problem that cannot be recognized by manually indexing and searching in the absence of a model
In order to achieve at least one of the above objects, the present invention further provides an AI weighing method including the steps of:
pre-establishing an image identification model and an image index of a weighing object;
establishing an image score model according to the image identification model and the image index;
acquiring a feature vector of a weighing object image, respectively inputting the feature vector into an image recognition model and an image index, and calculating a total similarity score of a target image according to the image recognition model and the image index;
and outputting sorting and weighing results according to the calculated similarity score.
According to a preferred embodiment of the present invention, the method for calculating the total similarity score comprises the following steps:
acquiring a characteristic vector of the weighted image, and acquiring a retrieval result N according to the index;
calculating the distance d of the characteristic vector of the target image according to the retrieval resulti；
Calculating the similarity score of a single retrieval result:
calculating similarity scores of all retrieval results:
inputting the feature vector of the weight image into the image recognition model to output a similarity score C;
obtaining a total similarity score: s ═ C + α score, where α is the adjustment weight.
According to another preferred embodiment of the present invention, the index construction method comprises: and carrying out granularity classification according to the attributes of the weights, and constructing an image index according to the feature vector generated by the image recognition model.
According to another preferred embodiment of the present invention, the training method of the image recognition model comprises: and setting a label on the weighing image in advance, pre-cutting according to the scale plate size of the weighing scale, obtaining the pre-cut weighing image, dividing the pre-cut weighing image into a training set and a testing set, and training the image recognition model.
According to another preferred embodiment of the present invention, a random rotation image, a shear image and a gaussian noise image are added to the training set to increase the robustness of the image recognition.
According to another preferred embodiment of the present invention, the training method of the image recognition model comprises: and carrying out pixel reduction on the pre-cut image, inputting the reduced image into the image recognition model, carrying out model training by adopting an Adam optimization algorithm, and adjusting the learning rate by adopting cosine learning rate decay.
According to another preferred embodiment of the present invention, the labels comprise a primary label and a secondary label, and are trained using F1 score loss (F1 fractional loss function), and a channel loss is added for regularization of the model.
According to another preferred embodiment of the invention, the established index is retrieved by adopting an HNSW indexing algorithm, and images with similar characteristics are retrieved according to the index.
In order to achieve at least one of the above objects, the present invention further provides an AI electronic scale system using the above AI weighing method.
The invention further provides a computer readable storage medium which stores and applies the AI electronic scale system.
Drawings
Fig. 1 shows a schematic flow diagram of an AI weighing method according to the invention.
FIG. 2 is a schematic view showing the processing flow of the AI electronic scale according to the present invention.
FIG. 3 is a schematic diagram showing the AI electronic scale correction process according to the present invention.
FIG. 4 is a schematic diagram showing the AI electronic scale recognition process of the present invention
FIG. 5 is a schematic view showing a detailed process flow of the AI electronic scale system according to the present invention.
Detailed Description
The following description is presented to disclose the invention so as to enable any person skilled in the art to practice the invention. The preferred embodiments in the following description are given by way of example only, and other obvious variations will occur to those skilled in the art. The basic principles of the invention, as defined in the following description, may be applied to other embodiments, variations, modifications, equivalents, and other technical solutions without departing from the spirit and scope of the invention.
It is understood that the terms "a" and "an" should be interpreted as meaning that a number of one element or element is one in one embodiment, while a number of other elements is one in another embodiment, and the terms "a" and "an" should not be interpreted as limiting the number.
Referring to fig. 1-5, the present invention discloses an AI electronic scale system and a weighing method, wherein a camera is required to be installed on an AI electronic scale, the camera can acquire image data of a weighed object in front of a weighing platform, and form a data stream through real-time shooting, the data stream is input into a machine learning model for recognition, the machine learning model performs similarity calculation according to a training set of the machine learning model and the acquired image of the weighed object, and displays a corresponding commodity list according to a similarity result.
In addition, in order to realize the application of the AI electronic weighing in the off-line state, the invention stores the AI electronic weighing in a local memory of the AI electronic scale by establishing an index mode on the basis of a machine learning model, thereby realizing the accurate search in the off-line state without manual query and improving the weighing efficiency.
Specifically, in a preferred embodiment of the present invention, an image recognition model of the weighing object is built at the cloud, which may be, but not limited to, a MobileNetv2 model, and the image recognition model is built at the cloud and trained, and then indexed into the AI electronic scale system for recognizing the weighing object image. And after the camera acquires the image information of the weighing object image, inputting the weighing object image information into the image recognition model, generating an image vector of the weighing object image by the image recognition model, and searching the target commodity image according to the image vector. In order to better illustrate the invention, the invention takes the MobileNetv2 model as an example, and explains the training and recognition mechanism of the image recognition model: and acquiring related image data in advance, and preprocessing the image data. The preprocessing process includes, but is not limited to, setting a label for the image data, and cutting the image data, and the manner of setting the label includes, but is not limited to, setting a commodity number and a commodity name corresponding to the image. Because the original image data acquired by the camera through shooting may include some irrelevant information, such as a weighing platform edge, a weighing shaft, and the like, the original image data needs to be cut, and the image edge of the element is cut according to the size of the weighing platform to obtain the acquired image data. The cropped image data is further divided into a training set and a test set at 7 for training the MobileNetv2 model. It should be noted that, in order to increase the robustness of the image recognition model, the present invention may be trained to add image enhancement functions including, but not limited to, random transformation, shear, and gaussian noise. The images in the training set are further reduced to 224 × 224, and the reduced images are input to the MobileNetv2 model for training.
In the training process of the MobileNetv2 model, an Adam optimization algorithm is adopted, and a cosine learning rate decay method is used for adjusting the learning rate, so that the MobileNetv2 model can be converged faster. Furthermore, the invention sets labels for each image, wherein the labels comprise a main label and an auxiliary label, and the learning is carried out by adopting a multi-label learning mode. For example, the weight is called apple, the labels of the apple and the red Fuji can be apple and the accuracy of model output can be improved by adopting a multi-label learning mode. The invention preferably trains by F1 score loss (F1 fractional loss function), and the training can be carried out by increasing channel loss (channel loss function) in the training process. The accuracy of identification can be greatly improved.
The invention establishes an image index on the basis of an image identification model, and establishes an image score model according to an image identification result and the image index, wherein the image score model needs to be fused with image identification and index to form the image score model, and the construction method of the image score model comprises the following steps:
the symmetrical weight A is only limited to be weighed, an image feature vector B is obtained according to the image recognition model, N retrieval results are retrieved through the established index, the feature vectors of the images corresponding to the N retrieval results are obtained according to the N retrieval results, and the distance d between the image feature vector corresponding to a single retrieval result and the feature vector B of the target weighing object A is calculatediWherein i represents a corresponding search result sequence, and i is more than or equal to 1 and less than or equal to N. Calculating the score s of a single search resulti:
Further calculating the total score of all the retrieval results of the weighed object A:
further calculating a score value C in the classifier of the image recognition model of the weighed object a, the total score value S of the weighed object a can be calculated as: s ═ C + α score; and alpha is the adjusting weight, the calculated total score of the weighing object A is sorted from large to small, and an index result corresponding to the top 10 of the score is output. Because the score model simultaneously considers the characteristics of the image recognition model and the index retrieval characteristics, the problem that a single recognition result or a single retrieval result has large deviation can be effectively avoided.
In one preferred embodiment of the present invention, if the image recognition model cannot recognize the image, the image feature vector corresponding to the selected search result is extracted directly through manual search, the feature vector and the photographed target weighing non-image are stored in the index, and then the feature vector and the photographed target weighing non-image are output according to the stored index result, and still the target weighing non-image can be sorted according to the score model, and the sorted score is output. That is, when the image recognition model fails, the total score S is α score, so that the problem that the model does not exist and cannot be recognized can be avoided.
It should be noted that the image recognition model in the present invention may be established at a cloud end, and a corresponding index may be established locally in the AI electronic scale system, where the interaction process between the AI electronic scale system and the cloud end includes:
after the user selects the finished product, storing the serial number of the product and the image vector of the model into an index;
after the selection is finished, the system uploads the selected commodity and the image address to the cloud end;
the cloud end pulls the reported log;
the cloud downloads the images in the log;
preprocessing the downloaded image;
and performing model training and index generation based on the downloaded images and the selected commodities in the log.
It should be noted that, in the present invention, for the established index, HNSW (Hierarchical navigation Small World diagram) is used as an index algorithm for searching. Results can be retrieved from an index of 10 ten thousand data in 2ms on the Intel i5 CPU and around 10 ms on the scale side.
In another preferred embodiment of the present invention, the image score model may further introduce weighing object categories to perform score sorting, for example, classification may be performed according to fine granularity of fruits and vegetables, category categories of the weighing objects are set, and category sorting weights are set, the camera may directly obtain the weighing category scores after identifying the weighing objects, and the weighing category scores are fused into the score model according to the weight values of the category scores to obtain weighing object category results and map the weighing object category results to corresponding weighing object labels (IDs), and score fusion is performed according to the weighing labels to obtain results with classification sorting and sorting scores. And (5) visually displaying the score result corresponding to the commodity information of the top 10.
In particular, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication section, and/or installed from a removable medium. The computer program, when executed by a Central Processing Unit (CPU), performs the above-described functions defined in the method of the present application. It should be noted that the computer readable medium mentioned above in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wire segments, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless section, wire section, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
It will be understood by those skilled in the art that the embodiments of the present invention described above and illustrated in the drawings are given by way of example only and not by way of limitation, the objects of the invention having been fully and effectively achieved, the functional and structural principles of the present invention having been shown and described in the embodiments, and that various changes or modifications may be made in the embodiments of the present invention without departing from such principles.
Claims (10)
1. An AI weighing method, characterized in that the weighing method comprises the following steps:
pre-establishing an image identification model and an image index of a weighing object;
establishing an image score model according to the image identification model and the image index;
acquiring a feature vector of a weighing object image, respectively inputting the feature vector into an image recognition model and an image index, and calculating a total similarity score of a target image according to the image recognition model and the image index;
and outputting sorting and weighing results according to the calculated similarity score.
2. The AI weighing method according to claim 1, wherein the overall similarity score calculation method includes the steps of:
acquiring a characteristic vector of the weighted image, and acquiring a retrieval result N according to the index;
calculating the distance d of the characteristic vector of the target image according to the retrieval resulti；
Calculating the similarity score of a single retrieval result:
calculating similarity scores of all retrieval results:
inputting the feature vector of the weight image into the image recognition model to output a similarity score C;
obtaining a total similarity score: s ═ C + α score, where α is the adjustment weight.
3. The AI weighing method of claim 1, wherein the index construction method comprises: and carrying out granularity classification according to the attributes of the weights, and constructing an image index according to the feature vector generated by the image recognition model.
4. The AI weighing method of claim 1, wherein the training method of the image recognition model comprises: and setting a label on the weighing image in advance, pre-cutting according to the scale plate size of the weighing scale, obtaining the pre-cut weighing image, dividing the pre-cut weighing image into a training set and a testing set, and training the image recognition model.
5. The AI weighing method of claim 4, wherein images including randomly rotated images, shear images, and gaussian noise are added to the training set to increase robustness of the image recognition.
6. The AI weighing method of claim 4, wherein the trainer of the image recognition model comprises: and carrying out pixel reduction on the pre-cut image, inputting the reduced image into the image recognition model, carrying out model training by adopting an Adam optimization algorithm, and adjusting the learning rate by adopting cosine learning rate decay.
7. The AI weighing method of claim 4, wherein the labels include a primary label and a secondary label, and are trained using F1 score loss (F1 fractional loss function) and adding channel loss for regularization of the model.
8. The AI weighing method according to claim 1, characterized in that the established index is retrieved using HNSW indexing algorithm, and images of similar characteristics are retrieved according to the index.
9. An AI electronic scale system characterized in that it employs an AI weighing method according to any one of the preceding claims 1 to 8.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores and applies an AI electronic scale system of claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110583656.0A CN113177525A (en) | 2021-05-27 | 2021-05-27 | AI electronic scale system and weighing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110583656.0A CN113177525A (en) | 2021-05-27 | 2021-05-27 | AI electronic scale system and weighing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113177525A true CN113177525A (en) | 2021-07-27 |
Family
ID=76927265
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110583656.0A Pending CN113177525A (en) | 2021-05-27 | 2021-05-27 | AI electronic scale system and weighing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113177525A (en) |
Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107886062A (en) * | 2017-11-03 | 2018-04-06 | 北京达佳互联信息技术有限公司 | Image processing method, system and server |
| CN108596090A (en) * | 2018-04-24 | 2018-09-28 | 北京达佳互联信息技术有限公司 | Facial image critical point detection method, apparatus, computer equipment and storage medium |
| CN109345733A (en) * | 2018-09-07 | 2019-02-15 | 杭州物宜网络科技有限公司 | The pricing method and system of intelligent scale |
| CN110189336A (en) * | 2019-05-30 | 2019-08-30 | 上海极链网络科技有限公司 | Image generating method, system, server and storage medium |
| CN110222709A (en) * | 2019-04-29 | 2019-09-10 | 上海暖哇科技有限公司 | A kind of multi-tag intelligence marking method and system |
| CN110458107A (en) * | 2019-08-13 | 2019-11-15 | 北京百度网讯科技有限公司 | Method and apparatus for image recognition |
| US20190354609A1 (en) * | 2018-05-21 | 2019-11-21 | Microsoft Technology Licensing, Llc | System and method for attribute-based visual search over a computer communication network |
| CN110704650A (en) * | 2019-09-29 | 2020-01-17 | 携程计算机技术（上海）有限公司 | OTA picture tag identification method, electronic device and medium |
| CN110807465A (en) * | 2019-11-05 | 2020-02-18 | 北京邮电大学 | Fine-grained image identification method based on channel loss function |
| CN110836717A (en) * | 2019-11-25 | 2020-02-25 | 湖北经济学院 | Financial service-oriented intelligent fruit and vegetable identification and pricing system |
| CN110942035A (en) * | 2019-11-28 | 2020-03-31 | 浙江由由科技有限公司 | Method, system, device and storage medium for acquiring commodity information |
| CN111090768A (en) * | 2019-12-17 | 2020-05-01 | 杭州深绘智能科技有限公司 | Similar image retrieval system and method based on deep convolutional neural network |
| CN111126514A (en) * | 2020-03-30 | 2020-05-08 | 同盾控股有限公司 | Image multi-label classification method, device, equipment and medium |
| CN111159456A (en) * | 2019-12-30 | 2020-05-15 | 云南大学 | Multi-scale clothing retrieval method and system based on deep learning and traditional features |
| CN111462093A (en) * | 2020-04-02 | 2020-07-28 | 北京小白世纪网络科技有限公司 | Method for classifying diseases based on fundus images |
| CN111625667A (en) * | 2020-05-18 | 2020-09-04 | 北京工商大学 | Three-dimensional model cross-domain retrieval method and system based on complex background image |
| CN111814614A (en) * | 2020-06-28 | 2020-10-23 | 袁精侠 | Intelligent object-identifying electronic scale weighing method and system |
| CN111860670A (en) * | 2020-07-28 | 2020-10-30 | 平安科技（深圳）有限公司 | Domain adaptive model training method, image detection method, device, equipment and medium |
| CN112016448A (en) * | 2020-08-27 | 2020-12-01 | 上海聚水潭网络科技有限公司 | System and method for image recognition of stored goods |
| CN112665698A (en) * | 2020-12-15 | 2021-04-16 | 重庆电子工程职业学院 | Intelligent electronic scale |
-
2021
- 2021-05-27 CN CN202110583656.0A patent/CN113177525A/en active Pending
Patent Citations (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107886062A (en) * | 2017-11-03 | 2018-04-06 | 北京达佳互联信息技术有限公司 | Image processing method, system and server |
| CN108596090A (en) * | 2018-04-24 | 2018-09-28 | 北京达佳互联信息技术有限公司 | Facial image critical point detection method, apparatus, computer equipment and storage medium |
| US20190354609A1 (en) * | 2018-05-21 | 2019-11-21 | Microsoft Technology Licensing, Llc | System and method for attribute-based visual search over a computer communication network |
| CN109345733A (en) * | 2018-09-07 | 2019-02-15 | 杭州物宜网络科技有限公司 | The pricing method and system of intelligent scale |
| CN110222709A (en) * | 2019-04-29 | 2019-09-10 | 上海暖哇科技有限公司 | A kind of multi-tag intelligence marking method and system |
| CN110189336A (en) * | 2019-05-30 | 2019-08-30 | 上海极链网络科技有限公司 | Image generating method, system, server and storage medium |
| CN110458107A (en) * | 2019-08-13 | 2019-11-15 | 北京百度网讯科技有限公司 | Method and apparatus for image recognition |
| CN110704650A (en) * | 2019-09-29 | 2020-01-17 | 携程计算机技术（上海）有限公司 | OTA picture tag identification method, electronic device and medium |
| CN110807465A (en) * | 2019-11-05 | 2020-02-18 | 北京邮电大学 | Fine-grained image identification method based on channel loss function |
| CN110836717A (en) * | 2019-11-25 | 2020-02-25 | 湖北经济学院 | Financial service-oriented intelligent fruit and vegetable identification and pricing system |
| CN110942035A (en) * | 2019-11-28 | 2020-03-31 | 浙江由由科技有限公司 | Method, system, device and storage medium for acquiring commodity information |
| CN111090768A (en) * | 2019-12-17 | 2020-05-01 | 杭州深绘智能科技有限公司 | Similar image retrieval system and method based on deep convolutional neural network |
| CN111159456A (en) * | 2019-12-30 | 2020-05-15 | 云南大学 | Multi-scale clothing retrieval method and system based on deep learning and traditional features |
| CN111126514A (en) * | 2020-03-30 | 2020-05-08 | 同盾控股有限公司 | Image multi-label classification method, device, equipment and medium |
| CN111462093A (en) * | 2020-04-02 | 2020-07-28 | 北京小白世纪网络科技有限公司 | Method for classifying diseases based on fundus images |
| CN111625667A (en) * | 2020-05-18 | 2020-09-04 | 北京工商大学 | Three-dimensional model cross-domain retrieval method and system based on complex background image |
| CN111814614A (en) * | 2020-06-28 | 2020-10-23 | 袁精侠 | Intelligent object-identifying electronic scale weighing method and system |
| CN111860670A (en) * | 2020-07-28 | 2020-10-30 | 平安科技（深圳）有限公司 | Domain adaptive model training method, image detection method, device, equipment and medium |
| CN112016448A (en) * | 2020-08-27 | 2020-12-01 | 上海聚水潭网络科技有限公司 | System and method for image recognition of stored goods |
| CN112665698A (en) * | 2020-12-15 | 2021-04-16 | 重庆电子工程职业学院 | Intelligent electronic scale |
Non-Patent Citations (1)
| Title |
|---|
| 迈克·贝尼科: "深度学习快速实践 基于TensorFlow和Keras的深度神经网络优化与训练" * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110705294B (en) | Named entity recognition model training method, named entity recognition method and named entity recognition device | |
| CN107944020B (en) | Face image searching method and device, computer device and storage medium | |
| US20180342077A1 (en) | Teacher data generation apparatus and method, and object detection system | |
| US9218364B1 (en) | Monitoring an any-image labeling engine | |
| US9349076B1 (en) | Template-based target object detection in an image | |
| CN108470172B (en) | Text information identification method and device | |
| WO2019089578A1 (en) | Font identification from imagery | |
| US9037600B1 (en) | Any-image labeling engine | |
| CN111931664A (en) | Mixed note image processing method and device, computer equipment and storage medium | |
| CN109165645A (en) | A kind of image processing method, device and relevant device | |
| CA3066029A1 (en) | Image feature acquisition | |
| CN109902285B (en) | Corpus classification method, corpus classification device, computer equipment and storage medium | |
| CN111860510B (en) | X-ray image target detection method and device | |
| CN111126514A (en) | Image multi-label classification method, device, equipment and medium | |
| CN108460114B (en) | Image retrieval method based on hierarchical attention model | |
| CN113963147B (en) | Key information extraction method and system based on semantic segmentation | |
| CN111461101B (en) | Method, device, equipment and storage medium for identifying work clothes mark | |
| CN105989001B (en) | Image search method and device, image search system | |
| KR20220125719A (en) | Method and equipment for training target detection model, method and equipment for detection of target object, electronic equipment, storage medium and computer program | |
| Hussain et al. | A simple and efficient deep learning-based framework for automatic fruit recognition | |
| CN114332473A (en) | Object detection method, object detection device, computer equipment, storage medium and program product | |
| CN111738199A (en) | Image information verification method, image information verification device, image information verification computing device and medium | |
| CN111898418A (en) | Human body abnormal behavior detection method based on T-TINY-YOLO network | |
| CN112270671B (en) | Image detection method, device, electronic equipment and storage medium | |
| CN113591850A (en) | Two-stage trademark detection method based on computer vision robustness target detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210727 |
|
| RJ01 | Rejection of invention patent application after publication |