WO2022062027A1 - 酒品定位方法、酒品信息管理方法、装置、设备及存储介质 - Google Patents

酒品定位方法、酒品信息管理方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2022062027A1
WO2022062027A1 PCT/CN2020/123637 CN2020123637W WO2022062027A1 WO 2022062027 A1 WO2022062027 A1 WO 2022062027A1 CN 2020123637 W CN2020123637 W CN 2020123637W WO 2022062027 A1 WO2022062027 A1 WO 2022062027A1
Authority
WO
WIPO (PCT)
Prior art keywords
wine
image
target
preset
label
Prior art date
Application number
PCT/CN2020/123637
Other languages
English (en)
French (fr)
Inventor
冯家禧
Original Assignee
冯家禧
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 冯家禧 filed Critical 冯家禧
Priority to GB2304282.3A priority Critical patent/GB2613753A/en
Priority to JP2023543250A priority patent/JP7502570B2/ja
Publication of WO2022062027A1 publication Critical patent/WO2022062027A1/zh
Priority to US18/127,006 priority patent/US20230237825A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/19—Recognition using electronic means
    • G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/12—Hotels or restaurants
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/19—Recognition using electronic means
    • G06V30/19007—Matching; Proximity measures
    • G06V30/19013—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/24—Indexing scheme for image data processing or generation, in general involving graphical user interfaces [GUIs]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30241—Trajectory

Definitions

  • the present application relates to the technical field of wine information management, and in particular, to a wine positioning method, a wine information management method, an apparatus, a computer device, and a computer-readable storage medium.
  • the storage of wine is generally stored in the wine cellar, the storage of wine requires specific environmental conditions for the wine cellar, and the management of the wine in the wine cellar, for example, for the storage of wine, especially for temperature and humidity
  • the requirements of the wine cellar are extremely demanding, and due to the functional limitations of the wine cellar, the wine stored in the wine cellar is mainly managed manually, and the wine in the wine cellar is managed automatically. , it is necessary to manage the wine according to the location of the wine in the wine cellar, that is, it is necessary to locate the wine in the wine cellar.
  • due to the inability to accurately and automatically locate the wine in the wine cellar resulting in The automated management of wine is less efficient.
  • the present application provides a wine positioning method, wine information management method, device, computer equipment and computer-readable storage medium, which can solve the problem of low positioning accuracy of wine in a wine cellar in the traditional technology, thereby solving In the traditional technology, the automatic management of wine through the wine cellar is inefficient.
  • the present application provides a wine positioning method, the method includes: based on a preset camera in a wine cellar, acquiring a wine image corresponding to a target wine collected by the preset camera; based on OCR character recognition
  • the preset wine label recognition method combined with deep learning recognition, recognizes the wine image to obtain the wine label corresponding to the wine image; obtains the preset collection position corresponding to the camera, and uses The preset collection position is used as the current position corresponding to the target wine; the wine label and the current position are used to describe the position corresponding to the target wine, so as to locate the target wine.
  • the present application provides a wine information management method applied to a server, the method comprising: acquiring wine information, where the wine information includes a wine image; The wine product corresponding to the product image is positioned to obtain the target position corresponding to the wine product; based on the target position, the wine product information is managed.
  • the present application provides a wine information management method applied to a terminal, the method includes: in response to a user operation, sending a request for obtaining wine information to a preset server, so that the preset server the wine information acquisition request, obtain the target wine information corresponding to the wine information acquisition request, and return the target wine information to the terminal, wherein the target wine information includes The wine product information associated with the corresponding wine product location, the wine product location is the location of the wine product obtained by locating the wine product according to the wine product positioning method; receiving the information sent by the preset server; and displaying the target wine information, so that the user can obtain the target wine information.
  • the present application also provides a wine positioning device, comprising: a first acquisition unit configured to obtain, based on a preset camera in the wine cellar, a wine image corresponding to a target wine collected by the preset camera Recognition unit for identifying the wine image based on the preset wine label recognition method combining OCR character recognition and deep learning recognition to obtain the wine label corresponding to the wine image; the second acquisition unit , used to obtain the preset collection position corresponding to the camera, and use the preset collection position as the current position corresponding to the target wine; the first positioning unit is used to use the wine label and the The current position describes the position corresponding to the target wine, so as to locate the target wine.
  • the present application also provides a wine information management device applied to a server, comprising: a third acquiring unit for acquiring wine information, the wine information including a wine image; a second positioning unit, is used to locate the wine corresponding to the wine image according to the wine positioning method, so as to obtain the target position corresponding to the wine; the management unit is used to locate the wine based on the target position. information management.
  • the present application also provides a wine information management device for an application terminal, comprising: a sending unit for sending a request for obtaining wine information to a preset server in response to a user operation, so that the preset server According to the wine information acquisition request, acquire the target wine information corresponding to the wine information acquisition request, and return the target wine information to the terminal, wherein the target wine information includes wine-based information
  • the wine product information associated with the wine product position corresponding to the product, the wine product position is the location of the wine product obtained by locating the wine product according to the wine product positioning method;
  • the receiving unit is used to receive the wine product.
  • the application also provides a kind of computer equipment, it comprises a memory and a processor, and a computer program is stored on the memory, and the processor executes the steps of the wine positioning method when the computer program is executed, Or execute the steps of the wine information management method applied to the server, or execute the steps of the wine information management method applied to the terminal.
  • the present application also provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the processor realizes the location of the wine product.
  • the present application provides a wine positioning method, wine information management method, device, computer equipment and computer-readable storage medium.
  • the present application obtains the wine image corresponding to the target wine collected by the preset camera based on the preset camera in the wine cellar, and the preset wine label recognition method based on the combination of OCR character recognition and deep learning recognition. Identify the wine image to obtain the wine label corresponding to the wine image; obtain the preset collection position corresponding to the camera, and use the preset collection position as the current position corresponding to the target wine , using the wine label and the current position to describe the position corresponding to the target wine, so as to locate the target wine.
  • the wine labels contained in the wine images can be accurately and quickly recognized, so that the wine labels contained in the wine images can be identified according to the identified
  • the wine label combined with the layout of the preset camera in the wine cellar, can realize the accurate positioning of the wine, so as to carry out accurate automatic management of the wine in the wine cellar, and avoid the wine cellar caused by the error of wine label recognition.
  • the confusion and errors in the automatic management of the wine in the wine cellar can improve the efficiency and quality of the automatic management of the wine in the wine cellar. Red wine is managed in real time, automatically and with high quality, thus improving the efficiency of automatic management of red wine.
  • Fig. 1 is an application environment diagram of the wine location method that the embodiment of this application provides;
  • Fig. 2 is a schematic flow chart of the wine positioning method provided by the embodiment of the application.
  • FIG. 3 is a schematic diagram of the first sub-process in the wine positioning method provided by the embodiment of the present application.
  • Fig. 4 is the second sub-flow schematic diagram of the wine product positioning method provided by the embodiment of this application.
  • FIG. 5 is a schematic flowchart of a wine information management method applied to a server side provided by an embodiment of the present application
  • FIG. 6 is a schematic flowchart of a method for managing wine information applied to a terminal provided by an embodiment of the present application
  • FIG. 7 is a schematic block diagram of a wine positioning device provided by an embodiment of the present application.
  • FIG. 8 is a schematic block diagram of a wine information management device applied to a server side provided by an embodiment of the present application
  • FIG. 9 is a schematic block diagram of a wine information management device applied to a terminal provided by an embodiment of the present application.
  • FIG. 10 is a schematic block diagram of a computer device provided by an embodiment of the present application.
  • FIG. 1 is an application environment diagram of the wine positioning method provided by the embodiment of the present application.
  • the application environment includes: (1) wine cellar. Several cameras are set up in different positions in the wine cellar.
  • the wine cellar of this example is provided with camera 1, camera 2, camera 3 and camera 4 to shoot the wine in the wine cellar, please continue Referring to the figure, as shown in Figure 1, in this example, it includes position 1, position 2 and position 3, camera 1 and camera 2 are used for shooting position 1, camera 3 is used for shooting position 2, and camera 4 is used for shooting position 3 , wine 1 wine can be wine in storage location 1, location 2 or location 3, etc.
  • red wine 1 wine can also be wine entering and leaving the wine cellar entrance and exit, for example, as shown in Figure 1, red wine 1 After entering the wine cellar, go through position 1 and position 2 to position 3, and then place it in position 3. On the basis of identifying the wine label corresponding to wine 1, combine the position 3 of wine 1 to determine the wine corresponding to wine 1. Information is managed automatically.
  • (2) Server It is used to receive the wine image corresponding to the wine 1 shot and uploaded by the camera, identify the wine label on the wine image, and combine the camera 1 and camera 2 according to the wine label recognition result corresponding to the identification of the wine label.
  • the preset collection positions corresponding to the camera 3 and the camera 4 especially the position 3 corresponding to the camera 4 where the red wine stays at the end, to manage the wine information of the wine corresponding to the wine 1.
  • Terminal It is used to receive the user's query, editing, deletion and other operations on the wine information corresponding to the red wine 1, and in response to the user's operation, obtain the wine information corresponding to the red wine 1 corresponding to the user's operation from the server.
  • the installed wine information management APP operates on wine information.
  • the operation process of each of the above subjects is as follows: 1) The camera in the wine cellar shoots the wine image corresponding to the wine in the wine cellar, and uploads the wine image to the server.
  • the camera 1 and the camera 2 in FIG. 1 shoot the wine 1.
  • the red wine and wine image 1 corresponding to the position 1 the camera 3 captures the red wine and wine image 2 corresponding to the red wine 1 in the position 2, and the camera 4 captures the red wine and the wine image 3 corresponding to the red wine 1 in the position 3, and the red wine is captured.
  • the wine image 1, the red wine image 2, and the red wine image 3 are uploaded to the server.
  • the server receives wine information, which includes wine images, such as receiving wine image 1, red wine image 2 or red wine image 3, and after obtaining the wine images, based on OCR text recognition and depth
  • a preset wine label recognition method combined with learning and recognition, identify the wine image to obtain the wine label corresponding to the wine image, and then obtain the preset collection position corresponding to the camera, which can be obtained through the camera.
  • the preset corresponding relationship with the position obtains the preset collection position corresponding to the wine captured by the camera, for example, the position 1 is corresponding to the shooting through the camera 1 and the camera 2, and the corresponding position is the shooting through the camera 3. It is position 2, and it can also identify the position image captured by the camera to identify the preset collection position corresponding to the captured wine.
  • the image The identification is to identify the position corresponding to the image as position 1, which may be to identify the position number corresponding to position 1, and use the preset collection position as the current position corresponding to the target wine. Use the wine label and the current position to describe the position corresponding to the target wine, so as to locate the target wine, and record, delete, and delete the wine information of the wine corresponding to the wine image. Editing and other management of wine information.
  • the target wine product information includes wine product information associated with the wine product position based on the wine product, and the wine product position is the wine product obtained by locating the wine product according to the wine product positioning method. location, return the target wine information to the terminal, receive the target wine information sent by the preset server, and display the target wine information, so that the user can obtain the information of the target wine. Describe the target wine information.
  • FIG. 2 is a schematic flowchart of a wine positioning method provided by an embodiment of the present application. As shown in Figure 2, the method includes the following steps S21-S24:
  • the wine when the wine is automatically managed in the wine cellar, it will involve locating the wine to obtain the location of the wine, and automatically manage the wine information corresponding to the wine according to the location of the wine.
  • Cameras can be set at different positions in the wine cellar according to the actual needs of wine shooting, so as to shoot the wine in the wine cellar through the camera.
  • the stored wine in the wine cellar and the wine taken out are photographed, and the camera uploads the wine image corresponding to the photographed wine to the computer equipment for wine positioning processing, so that the computer equipment obtains the target collected by the preset camera.
  • the wine image corresponding to the wine The wine image captured by the camera may be a picture or a video. If the camera captures a video, a video frame needs to be extracted from the video to obtain a wine image corresponding to the wine.
  • the wine image when performing wine label recognition on the wine image, the wine image is recognized based on a preset wine label recognition method combining OCR character recognition and deep learning recognition, so as to obtain the The wine label corresponding to the wine image, because OCR text recognition can quickly and accurately identify the text of the standard characters contained in the image, and deep learning recognition can accurately recognize the image, so based on OCR text recognition and deep learning recognition.
  • the combined preset wine label identification method can accurately and quickly identify the wine label contained in the wine image, and improve the accuracy and recognition efficiency of the wine label included in the wine image.
  • the wine cellar when automatically positioning the wine in the wine cellar, it is necessary to identify the wine label to which the wine belongs, that is, what wine the wine is, and then automatically locate the wine in combination with the location of the wine in the wine cellar
  • the obtained position corresponding to the wine that is, the position of the wine
  • the position of the wine in the wine cellar that is, the preset collection position corresponding to the camera
  • the preset collection position corresponding to the camera in the wine cellar It can be obtained through the fixed preset collection position corresponding to the camera in the wine cellar, that is, the preset collection position is fixed in advance for each camera, and the corresponding relationship between the camera and the fixed preset collection position is obtained according to the camera.
  • the preset collection position corresponding to the camera is acquired, and the preset collection position corresponding to the camera can be determined through the camera, thereby obtaining the preset collection position corresponding to the camera.
  • camera A is set at position A in the wine cellar, and the image of position B in the wine cellar is captured by camera A. If the wine image captured by camera A is received, it can be known that the wine corresponding to the wine image is in Location B in the cellar.
  • the wine product corresponding to the wine product image can be obtained in The position in the wine cellar where the wine image was taken, and the preset collection position is taken as the current position corresponding to the target wine.
  • the preset acquisition position corresponding to the camera can be acquired by means of image detection. That is, a camera is set corresponding to each wine placement position, and the placement position is photographed by the camera, for example, the position identification corresponding to the placement position is photographed, and the position identification can be a position number. , obtain the image corresponding to the placement position through the camera, and then perform image detection on the image corresponding to the placement position to identify the placement position. For example, camera A is set at position A in the wine cellar, and the image of position B where the wine is placed in the wine cellar is captured by camera A.
  • the image including position B is captured by the camera
  • the image containing the position B after receiving the image containing the position B captured by the camera A, perform image recognition on the image containing the position B to identify the position B, and you can know that the position of the wine corresponding to the wine image in the wine cellar is the position B , so as to obtain the preset collection position corresponding to the camera.
  • the wine label and the current location can be used to describe the wine The position corresponding to the target wine, so as to locate the target wine.
  • the wine label included in the wine image corresponding to a target wine is M
  • the position of the target wine corresponding to the wine image in the wine cellar is L
  • the position of the target wine in the wine cellar is L.
  • the positioning can be: wine M, L position in the wine cellar.
  • the wine labels contained in the wine images can be accurately and quickly recognized, so that the wine labels contained in the wine images can be identified according to the identified
  • the wine label combined with the position of the wine in the wine cellar, can realize the accurate positioning of the wine, so as to carry out accurate automatic management of the wine in the wine cellar, and avoid the error in the wine label recognition caused by the error in the wine cellar.
  • the confusion and errors in the automatic management of wine improve the efficiency and quality of automatic management of wine in the wine cellar.
  • the present application obtains the wine image corresponding to the target wine collected by the preset camera based on the preset camera in the wine cellar, and the preset wine label recognition method based on the combination of OCR character recognition and deep learning recognition. Identify the wine image to obtain the wine label corresponding to the wine image; obtain the preset collection position corresponding to the camera, and use the preset collection position as the current position corresponding to the target wine , using the wine label and the current position to describe the position corresponding to the target wine, so as to locate the target wine.
  • the wine labels contained in the wine images can be accurately and quickly recognized, so that the wine labels contained in the wine images can be identified according to the identified
  • the wine label combined with the layout of the preset camera in the wine cellar, can realize the accurate positioning of the wine, so as to carry out accurate automatic management of the wine in the wine cellar, and avoid the wine cellar caused by the error of wine label recognition.
  • the confusion and errors in the automatic management of the wine in the wine cellar can improve the efficiency and quality of the automatic management of the wine in the wine cellar. Red wine is managed in real time, automatically and with high quality, thus improving the efficiency of automatic management of red wine.
  • FIG. 3 is a schematic diagram of a first sub-flow of the wine positioning method provided by the embodiment of the present application.
  • the preset wine label recognition method based on the combination of OCR character recognition and deep learning recognition recognizes the wine image to obtain the corresponding wine image.
  • the steps of the wine label include the following steps S31-S33:
  • OCR OCR
  • English Optical Character Recognition
  • optical character recognition which is to convert the text in the document into a black and white dot matrix image file by optical means, and convert the text in the image into text format through recognition software.
  • the wine image in the wine cellar is collected by the camera set in the wine cellar, and the wine image is uploaded to the computer equipment for wine identification, so that the computer equipment obtains the information to be identified.
  • the computer device performs OCR recognition on the wine image according to the preset OCR recognition method, so as to extract the text contained in the wine image, and use the text as the target text. Since the OCR technology is good at recognizing standard text and other character-type content, it can accurately identify the text contained in the wine image. For example, when the wine label is automatically identified in the wine cellar to manage the wine product information, in the wine wine cellar, several cameras deployed in the wine cellar can take pictures of the wine product.
  • the wine image can be a single image captured by a camera (that is, a picture in a picture format), and the wine image is uploaded to the background wine and wine information server, or it can be a video collected by a camera, and the The video collected by the camera is uploaded to the background wine and wine information server, and then by extracting the image obtained from the video frame in the video, the background wine and wine information server performs OCR identification on the wine image according to the preset OCR identification method, In order to obtain the text contained in the wine image, the text is used as the target text corresponding to the wine label.
  • the step of obtaining the wine image also includes:
  • the initial wine product image is segmented to obtain a wine product image corresponding to the partial image including the preset wine label in the initial wine product image.
  • the wine label to be identified is also within the preset wine label range.
  • the obtained initial wine image can be segmented, so as to Intelligently segment other images of wine products such as environmental images other than wine products and wine bottles other than wine labels contained in the initial wine product image, and take the segmented partial images containing the preset wine labels as the target wine
  • the approximate position of the label image is determined, and the target wine label image is used as a wine product image for wine label identification. Since the image for identification is focused on the local image containing the wine label, especially in the complex environment containing many environmental factors such as the wine cellar, the wine image corresponding to the wine bottle can be accurately identified, so as to obtain the corresponding accurate image.
  • the image features can improve the accuracy of subsequent identification of wine labels, and compared with the identification of the entire initial wine image, the amount of data processing is also reduced, thereby improving the efficiency of identification of wine labels.
  • Deep learning English for Deep Learning, is to establish and simulate a neural network for analysis and learning of the human brain. It imitates the mechanism of the human brain to interpret data.
  • the models of deep learning include Feed forward neural networks (FF or FFNN) and perceptrons ( P), Boltzmann machines (BM), etc., the deep learning models that can identify wine labels better can be selected from many deep learning models through test samples.
  • Deep learning can obtain the image features contained in the wine image. Since deep learning is good at learning the training sample images and the labels corresponding to the training sample images, the pre-trained deep learning-based deep learning model is good at recognizing image features, so that the wine image can be accurately identified.
  • the image features included in the image feature, and the image feature is used as the target image feature corresponding to the wine image. For example, when automatically recognizing red wine labels for red wine information management, after obtaining the red wine images corresponding to the red wines, depth-recognizing the red wine images is performed according to a preset deep learning recognition method. Learning to recognize is to recognize the image features contained in the red wine image, and use the image features as the target image features corresponding to the red wine image.
  • the process of training the preset deep learning model corresponding to the preset deep learning identification method includes the following steps:
  • the wine label image pre-training sample includes the training wine label image and the corresponding training wine label image.
  • training wine label is photographed for the wine label corresponding to the wine product to be managed, especially the wine label image on the wine product is photographed from various angles, and the captured wine label image is labeled, so-called labeling , the wine label image to be photographed is associated with the wine label name corresponding to the wine label to describe that the wine label image corresponds to the wine label name, and the wine label name is the wine label label corresponding to the wine label image. .
  • the wine label image pre-training sample is input into the preset deep learning model, so that the preset deep learning model automatically learns according to the training wine label image and the training wine label label corresponding to the training wine label image, Learning the correlation between the training image features corresponding to the training wine label images and the training wine label labels, so as to complete the training of the deep learning model.
  • the preset deep learning model uses a wine label image test sample, where the wine label image test sample includes the test wine label image and the test wine label label corresponding to the test wine label image.
  • the trained preset deep learning model is tested with a wine label image test sample, the preset deep learning model identifies the test wine label image to obtain the test target wine label, and the test target wine label is Compare with the test wine label corresponding to the test wine label image, and judge whether the test target wine label is consistent with the test wine label corresponding to the test wine label image.
  • the test wine label corresponding to the test wine label image is consistent, indicating that the deep learning model has a relatively accurate recognition effect on the preset wine label recognition.
  • test target wine label and the test wine corresponding to the test wine label image are The labels are inconsistent, indicating that the deep learning model does not have a relatively accurate recognition effect on the preset wine label recognition, so as to judge whether the preset deep learning model achieves the expected goal, and if it reaches the expected goal, put the preset deep learning model into the production environment , if the expected goal is not achieved, re-train with more training samples, adjust the training samples, and adjust the preset deep learning model until the preset deep learning model finally meets the expected goal.
  • the wine label constitutes a preset wine label database. After extracting the text and image features contained in the wine image from the wine image, the text and the image feature can be combined according to the text and the image feature. Then, from the preset wine label database, the wine label that matches the text and the image features is selected as the target wine label, and the target wine label is used as the wine label corresponding to the wine image, so as to The wine label corresponding to the wine image is identified, and the wine information is automatically managed according to the identified wine label.
  • the red wine information is managed through the wine cellar, after the wine image is obtained, the text and image features in the red wine image are identified, and the character and image features are identified according to the text and the image features.
  • the red wine information is managed according to the identified red wine label.
  • Image recognition based on deep learning technology requires a large amount of calculation.
  • it will also be more efficient for wine label recognition.
  • it is difficult to accumulate a large enough wine label training data sample for the application environment of wine information management. Therefore, it is extremely difficult for computer equipment to accurately identify various wine labels.
  • the applicant of the embodiment of the present application faces a situation where there are huge differences in various wine labels in a complex environment and needs to be uniformly and accurately recognized. Based on the characteristics of wine labels consisting of text and images, OCR text recognition and deep learning image recognition are used.
  • a target wine label matching the text and the image features is selected from a preset wine label database, and the target wine label is selected.
  • the steps of serving as the wine label corresponding to the wine image include:
  • the preset wine labels matching the text are selected from the preset wine label database, and the preset wine labels are formed into a set to obtain a target wine label set;
  • the wine label with the highest matching degree with the image feature is selected from the target wine label set as the target wine label, and the target wine label is used as the wine corresponding to the wine image. mark.
  • a preset wine label matching the text is selected from the preset wine label database, and the A set of preset wine labels is formed to obtain the target wine label set, so as to make full use of the advantages of OCR for fast and accurate recognition of standard text, and narrow the scope of deep learning search wine labels to the target wine label set.
  • the wine label is a combination of text and images, or the recognized text is inaccurate, for example, the recognized text is similar, the same, or contained in other text combinations, it cannot be recognized only by
  • the output text can accurately determine the wine label corresponding to the wine image.
  • the wine label is used as the target wine label, and the target wine label is used as the wine label corresponding to the wine image, so as to identify the wine label corresponding to the wine image.
  • the pre-trained deep learning model is more accurate for image recognition, when the search scope of the deep learning model has been narrowed to the target wine label set, compared to directly searching from the preset wine label database to obtain the target wine Labels, which can greatly reduce the calculation amount of the deep learning model, and overcome the defect of low recognition accuracy in complex environments such as wine cellars when the OCR model recognizes text, so as to screen out the target wine label collection.
  • the target wine label can obtain an accurate wine label corresponding to the wine image, which improves the accuracy and efficiency of identifying the wine label corresponding to the wine image.
  • FIG. 4 is a schematic diagram of a second sub-flow of the wine positioning method provided by the embodiment of the present application.
  • the preset camera includes several cameras, and each camera captures wine images corresponding to wines in different positions, or each camera captures wines in the same position from different angles
  • the preset camera includes several cameras, and each camera captures wine images corresponding to wines in different positions, or each camera captures wines in the same position from different angles
  • the camera By setting the camera at the same position and from different angles to shoot the wine at the position, it can avoid inaccurate or unclear shooting of the wine due to the problem of the camera shooting angle, and try to improve the clarity of the wine at the position as much as possible.
  • Accurate wine image and then identify the wine image accurately, so as to obtain wine information such as wine label and wine location corresponding to the accurate wine.
  • the wine image corresponding to the wine can be photographed. Since there are multiple cameras in the wine cellar to photograph the wine, the wine image includes several images, the acquisition of the preset collection position corresponding to the cameras, and the preset collection position as the The steps of the current position corresponding to the target wine include:
  • the wine image includes several images, and the several images are acquired by several cameras pre-installed in the preset wine cellar, and the several cameras can be installed in different preset positions in the wine cellar , when the wine moves in the wine cellar, the target wine can be photographed from different positions, or the same target wine can be photographed from different angles of the same position.
  • multiple cameras are generally set in the wine cellar.
  • a plurality of cameras are installed in the wine cellar, and several of the cameras are installed in different preset positions in the wine cellar to shoot the target red wine from different positions or different angles of the same position.
  • the wine images include the shooting time of the wine images and the preset camera identifier corresponding to the camera that uploaded the wine images, and the preset It is assumed that the camera is preset at a preset fixed position to capture the wine image corresponding to the preset collection position.
  • the wine product image obtained by shooting to identify the wine label of the target wine product, and then obtain the preset collection position corresponding to the wine product image (that is, to obtain the wine product at that time). location), according to the identified wine label, the preset collection position corresponding to the obtained wine image and the shooting sequence of the wine image (that is, the shooting time sequence), all the preset The collection positions are sorted according to the shooting sequence of the corresponding wine images, so as to obtain a preset collection position sorting queue, so as to obtain the movement trajectory corresponding to the wine, and the above process is iterated until the wine is The position is no longer changed, and the last position is determined as the final position of the wine, so that the last preset collection position at the last position in the collection position sorting queue is screened out, and the last preset collection position is used as The current position corresponding to the target wine product realizes the positioning of the wine product. Based on the positioning of the wine, the wine information is then processed, so as to realize the automatic management of the target
  • Table 1 is an example of sorting all the wine images. If there are five cameras ABCDE, camera A ranks In the 1st position, the B camera is in the 2nd position, the C camera is in the 3rd position, the D camera is in the 4th position, and the E camera is in the 5th position. Connect the different positions to form a corresponding path. , the path corresponding to the movement of the object is the movement track. If the wine passes through the 12345 position in turn, the movement track of the wine can be obtained. If the wine stays at position 5 and no longer moves, the end point of the wine movement can be determined.
  • 12345 is a path
  • 134 is a path
  • 145 can be a path
  • the specific path can be set according to the layout of the wine cellar.
  • the complete path can be completed according to the known partial path, for example, if there is only 1234 unique paths from position 1 to position 4,
  • the movement trajectory corresponding to the wine is identified as 134 or 14 in the movement trajectory of , and it can also be determined according to 134 or 14 that the complete movement trajectory of the wine should be 1234.
  • Form 1 is shown below:
  • P ATA1 is used to describe the wine image of camera A at TA1 time
  • P ATA2 is used to describe the wine image of camera A at TA2 time
  • Image camera B contains no images of wine labels due to occlusion
  • camera C contains 2 images of wines that are identified to contain the same wine label
  • camera D contains 1 image of wine that contains the same wine label
  • camera E contains Five wine images containing the same wine label
  • the above wine images are the same wine, among which, camera A corresponds to position 1, camera B corresponds to position 2, camera C corresponds to position 3, camera D corresponds to position 4, and camera E corresponds to position 4.
  • the camera corresponds to position 5.
  • the movement trajectory of the wine is 1345, or, if the path is unique, although the B camera does not contain the image of the wine, since the position from 1 to 345 must pass through the B camera, the movement trajectory of the wine is 12345, since the position of the wine label that was finally recognized is at the No. 5 position corresponding to the E camera, and the preset time has passed, other cameras have not recognized the wine label, so it can be seen that the last position of the wine corresponding to the wine label is determined at The No.
  • the unique position of the wine cabinet can more clearly determine the position of the wine cabinet corresponding to the position of the red wine, thereby realizing the positioning of the red wine.
  • FIG. 5 is a schematic flowchart of a wine information management method applied to a server side provided by an embodiment of the present application. As shown in Figure 5, the method includes the following steps S51-S53:
  • the server automatically manages the wine product information corresponding to the wine product, it not only needs to obtain the wine product image, but also uses the wine product positioning method described in each of the above embodiments to perform automatic management on the wine product corresponding to the wine product image. Positioning, to obtain the wine product position corresponding to the wine product, and to obtain other aspects of the wine product such as the wine product year, wine product origin, production date and the entry and exit time of the wine product corresponding to the wine product After identifying the wine label corresponding to the wine image and the wine location corresponding to the wine, the wine information is added and modified based on the wine label and the wine location.
  • FIG. 6 is a schematic flowchart of a wine information management method applied to a terminal provided by an embodiment of the present application. As shown in Figure 6, the method includes the following steps S61-S63:
  • S61 in response to a user operation, send a wine information acquisition request to a preset server, so that the preset server acquires the target wine information corresponding to the wine information acquisition request according to the wine information acquisition request, and returning the target wine information to the terminal, wherein the target wine information includes wine information associated with the wine position corresponding to the wine, and the wine position is based on the above embodiments The location of the wine obtained by locating the wine by the described wine positioning method;
  • S63 Display the target wine information, so that the user can obtain the target wine information.
  • the terminal responds to the user's operation and sends a request for obtaining wine information to the preset server, so that the preset server can be based on the The wine information acquisition request, obtains the target wine information corresponding to the wine information acquisition request, and the target wine information includes the wine information associated with the wine position corresponding to the wine, and the wine information
  • the location is the location of the wine obtained by locating the wine according to the wine positioning method described in the above embodiments, and the target wine information is returned to the terminal, and the terminal receives The target wine product information sent by the preset server is displayed, so that the user can obtain the target wine product information, because the accuracy and efficiency of wine positioning and identification are improved.
  • FIG. 7 is a schematic block diagram of a wine label positioning device provided by an embodiment of the present application.
  • an embodiment of the present application further provides a wine label positioning device.
  • the wine label locating device includes a unit for executing the above-mentioned wine label locating method, and the wine label locating device can be configured in a computer device such as a server.
  • the wine label positioning device 70 includes a first obtaining unit 71 , an identifying unit 72 , a second obtaining unit 73 and a first positioning unit 74 .
  • the first acquisition unit 71 is configured to acquire, based on the preset camera in the wine cellar, the wine image corresponding to the target wine collected by the preset camera;
  • the identification unit 72 is used to identify the wine image based on the preset wine label recognition method combining OCR character recognition and deep learning recognition to obtain the wine label corresponding to the wine image;
  • the second acquiring unit 73 is configured to acquire the preset collection position corresponding to the camera, and use the preset collection position as the current position corresponding to the target wine;
  • the first positioning unit 74 is configured to describe the position corresponding to the target wine by using the wine label and the current position, so as to locate the target wine.
  • the identifying unit 72 includes:
  • the first identification subunit is used to perform OCR identification on the wine image according to a preset OCR identification method, so as to obtain the text contained in the wine image;
  • the second identification subunit is used to perform deep learning identification on the wine image according to a preset deep learning identification method, so as to obtain the image features contained in the wine image;
  • the first screening subunit is used to screen out the target wine label that matches the character and the image feature from the preset wine label database according to the character and the image feature, and select the target wine label. as the wine label corresponding to the wine image.
  • the second obtaining unit 73 includes:
  • a first acquisition subunit configured to acquire the preset collection positions corresponding to all the wine images, so as to obtain all the preset collection positions
  • a sorting subunit configured to sort all the preset collection positions according to the shooting sequence of the corresponding wine images, so as to obtain a collection position sorting queue
  • the second screening subunit is configured to filter out the last preset collection position in the collection position sorting queue, and use the last preset collection position as the current position corresponding to the target wine.
  • FIG. 8 is a schematic block diagram of an apparatus for managing wine information applied to a server according to an embodiment of the present application.
  • an embodiment of the present application further provides a wine information management device applied to the server.
  • the wine information management device 80 applied to the server includes a unit for executing the above-mentioned wine information management method applied to the server, and the wine information management device 80 can be configured on a computer such as a server in the device.
  • the wine information management device 80 includes a third acquisition unit 81 , a second positioning unit 82 and a management unit 83 .
  • the third obtaining unit 81 is used for obtaining wine product information
  • the wine product information includes wine product images
  • the second positioning unit 82 is used for, according to the wine product positioning methods described in the above embodiments, to The wine product corresponding to the wine product image is located to obtain the target position corresponding to the wine product
  • the management unit 83 is configured to manage the wine product information based on the target position.
  • FIG. 9 is a schematic block diagram of an apparatus for managing wine information applied to a terminal according to an embodiment of the present application.
  • an embodiment of the present application further provides a wine information management device applied to a terminal.
  • the wine information management device 90 applied to the terminal includes a unit for executing the above-mentioned wine information management method applied to the terminal, and the wine information management device 90 can be configured on a smart phone or the like in the terminal device.
  • the wine information management device 90 includes a sending unit 91 , a receiving unit 92 and a display unit 93 .
  • the sending unit 91 is configured to, in response to a user operation, send a wine information acquisition request to a preset server, so that the preset server acquires the corresponding wine information acquisition request according to the wine information acquisition request.
  • target wine product information and return the target wine product information to the terminal, wherein the target wine product information includes wine product information associated with the wine product location based on the wine product, the wine product location is the location of the wine obtained by locating the wine according to the wine positioning method described in the above embodiments;
  • the receiving unit 92 is configured to receive the target sent by the preset server Wine product information;
  • the display unit 93 is configured to display the target wine product information, so that the user can obtain the target wine product information.
  • the embodiment of the present application also provides a wine information management system, the system includes a wine cellar, a terminal and a server; wherein, the wine cellar is provided with a wine cabinet, the wine cabinet is used for placing wine products, and the wine cellar is preset A camera is provided at the location, so that the wine image corresponding to the wine is captured by the camera, and the wine image is uploaded to the server; the server is used to execute the applications described in the above embodiments The steps of the wine information management method on the server side; the terminal is configured to execute the steps of the wine information management method applied to the terminal described in each of the above embodiments.
  • each unit in the above-mentioned wine locating device, the above-mentioned wine information management device applied to the server, or the above-mentioned wine information management device applied to the terminal are only used for illustration.
  • Divide the wine product positioning device, the above-mentioned wine product information management device applied to the server side or the above-mentioned wine product information management device applied to the terminal into different units as required, and the wine product positioning device, the above-mentioned wine product information management device applied to the server end can also be divided into different units.
  • Each unit in the information management device or the above-mentioned wine information management device applied to the terminal adopts different connection sequences and modes to complete the above-mentioned wine positioning device, the above-mentioned wine information management device applied to the server side or the above-mentioned wine applied to the terminal. All or part of the functions of the product information management device.
  • the above-mentioned wine product positioning device, the above-mentioned wine product information management device applied to the server side or the above-mentioned wine product information management device applied to the terminal can be respectively implemented as a form of a computer program, and the computer program can be respectively shown in Figure 10. run on computer equipment.
  • FIG. 10 is a schematic block diagram of a computer device provided by an embodiment of the present application.
  • the computer device 500 may be a computer device such as a desktop computer or a server, or may be a component or component in other devices.
  • the computer device 500 includes a processor 502, a memory and a network interface 505 connected through a system bus 501, wherein the memory may include a non-volatile storage medium 503 and an internal memory 504, and the memory may also be a volatile computer Readable storage medium.
  • the nonvolatile storage medium 503 can store an operating system 5031 and a computer program 5032 .
  • the computer program 5032 When executed, it can cause the processor 502 to execute the above-mentioned wine positioning method, the above-mentioned wine information management method applied to the server, or the above-mentioned wine information management method applied to the terminal.
  • the processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .
  • the internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503.
  • the processor 502 can execute the above-mentioned method for recognizing a wine label, the above-mentioned application for The wine information management method on the server side or the above-mentioned wine information management method applied to the terminal.
  • the network interface 505 is used for network communication with other devices.
  • the specific computer device 500 may include more or fewer components than shown, or combine certain components, or have a different arrangement of components.
  • the computer device may only include a memory and a processor. In such an embodiment, the structures and functions of the memory and the processor are the same as those of the embodiment shown in FIG. 10 , which will not be repeated here.
  • the processor 502 when the processor 502 is used to run the computer program 5032 stored in the memory, to realize the above-mentioned wine positioning method, the processor 502 performs the following steps: based on the preset camera in the wine cellar, obtain The wine image corresponding to the target wine collected by the preset camera; based on the preset wine label recognition method combining OCR character recognition and deep learning recognition, the wine image is recognized to obtain the wine image the wine label corresponding to the image; obtain the preset collection position corresponding to the camera, and use the preset collection position as the current position corresponding to the target wine; use the wine label and the description of the current position The position corresponding to the target wine is used to locate the target wine.
  • the processor 502 is implementing the preset wine label recognition method based on the combination of OCR character recognition and deep learning recognition, and recognizes the wine image to obtain the information of the wine image.
  • the following steps are specifically implemented:
  • the wine image is carried out OCR identification to obtain the text contained in the wine image
  • a target wine label matching the text and the image features is selected from the preset wine label database, and the target wine label is used as the wine image corresponding wine label.
  • the preset camera includes several cameras
  • the wine image includes several images
  • the processor 502 is implementing the acquisition of the preset collection position corresponding to the camera, and storing all the images.
  • the last preset collection position at the last position in the collection position sorting queue is screened out, and the last preset collection position is used as the current position corresponding to the target wine.
  • the processor 502 when the processor 502 is used to run the computer program 5032 stored in the memory to realize the above-mentioned wine information management method applied to the server side, the processor 502 performs the following steps: obtaining wine. wine product information, the wine product information includes wine product images; according to the wine product positioning methods described in the above embodiments, the wine product corresponding to the wine product image is located to obtain the wine product corresponding to the wine product The target location; based on the target location, the wine information is managed. .
  • the processor 502 executes the following steps: in response to User operation, send a wine information acquisition request to a preset server, so that the preset server acquires the target wine information corresponding to the wine information acquisition request according to the wine information acquisition request, and sends the The target wine product information is returned to the terminal, wherein the target wine product information includes wine product information associated with the wine product position corresponding to the wine product, and the wine product position is based on the wine products described in the above embodiments.
  • the location of the wine product obtained by locating the wine product by the wine product positioning method; receiving the target wine product information sent by the preset server; The user acquires the target wine information. .
  • the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like.
  • the present application also provides a computer-readable storage medium.
  • the computer-readable storage medium may be a non-volatile computer-readable storage medium, or a volatile computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed by the processor Make the processor execute the steps of the wine positioning method described in the above embodiments, or cause the processor to execute the wine information management method described in the above embodiments and applied to the server side, or make the processing
  • the device executes the above-mentioned wine information management method applied to the terminal described in the above embodiments.
  • the computer-readable storage medium may be an internal storage unit of the aforementioned device, such as a hard disk or a memory of the device.
  • the computer-readable storage medium may also be an external storage device of the device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card equipped on the device , Flash Card (Flash Card) and so on.
  • the computer-readable storage medium may also include both an internal storage unit of the device and an external storage device.
  • the storage medium is a physical, non-transitory storage medium, such as a U disk, a removable hard disk, a read-only memory (Read-Only Memory, ROM), a magnetic disk or an optical disk and other physical storage that can store computer programs. medium.
  • a physical, non-transitory storage medium such as a U disk, a removable hard disk, a read-only memory (Read-Only Memory, ROM), a magnetic disk or an optical disk and other physical storage that can store computer programs. medium.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the integrated unit if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a storage medium.
  • the technical solutions of the present application are essentially or part of contributions to the prior art, or all or part of the technical solutions can be embodied in the form of software products, and the computer software products are stored in a storage medium , including several instructions to cause an electronic device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • General Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Pathology (AREA)
  • Chemical & Material Sciences (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Analytical Chemistry (AREA)
  • Primary Health Care (AREA)
  • Immunology (AREA)
  • Biochemistry (AREA)
  • Image Analysis (AREA)

Abstract

一种酒品定位方法、酒品信息管理方法、装置、计算机设备及计算机可读存储介质。通过基于酒窖中预设摄像头，获取预设摄像头采集的目标酒品所对应的酒品图像(S21)，基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对酒品图像进行识别，以得到酒品图像所对应的酒标(S22)；获取摄像头所对应的预设采集位置，并将预设采集位置作为目标酒品所对应的当前位置(S23)，采用酒标及当前位置描述目标酒品所对应的位置，以对目标酒品进行定位(S24)，通过准确、快速的识别出酒品图像中包含的酒标，结合酒窖中预设摄像头的布局获取酒品的位置，能够实现对酒品的准确定位，以提高酒窖中酒品自动化管理的效率和质量。

Description

酒品定位方法、酒品信息管理方法、装置、设备及存储介质
本申请是以申请号为202011040743.3、申请日为2020年9月28日的中国专利申请为基础，并主张其优先权，该申请的全部内容在此作为整体引入本申请中。
技术领域
本申请涉及酒品信息管理技术领域，尤其涉及一种酒品定位方法、酒品信息管理方法、装置、计算机设备及计算机可读存储介质。
背景技术
酒类自古至今在社交中都扮演着重要角色，例如，葡萄酒在现代社交中扮演着越来越重要的角色。酒类的储存一般储存在酒窖中，酒类的储存，对于酒窖需要特定的环境条件，并且需要对酒窖中的酒品进行管理，例如，对于葡萄酒的储存，特别是对于温度和湿度的要求极为苛刻，并且传统的红酒酒窖由于酒窖的功能局限，对酒窖中储存的红酒酒品，主要是通过人工对酒品进行管理，而要对酒窖中的酒品进行自动化管理，需要根据酒窖中酒品所在的位置对酒品进行管理，即需要对酒窖中的酒品进行定位，而传统技术中，由于无法对酒窖中的酒品进行精确的自动定位，导致对酒品的自动化管理效率较低。
申请内容
本申请提供了一种酒品定位方法、酒品信息管理方法、装置、计算机设备及计算机可读存储介质，能够解决传统技术中对酒窖中的酒品定位准确性较低的问题，从而解决传统技术中通过酒窖对酒品进行自动化管理效率较低的问题。
第一方面，本申请提供了一种酒品定位方法，所述方法包括：基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
第二方面，本申请提供了一种应用于服务器端的酒品信息管理方法，所述方法包括：获取酒品信息，所述酒品信息包括酒品图像；根据酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；基于所述目标位置，对所述酒品信息进行管理。
第三方面，本申请提供了一种应用于终端的酒品信息管理方法，所述方法 包括：响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；接收所述预设服务器发送的所述目标酒品信息；将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
第四方面，本申请还提供了一种酒品定位装置，包括：第一获取单元，用于基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；识别单元，用于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；第二获取单元，用于获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；第一定位单元，用于采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
第五方面，本申请还提供了一种应用于服务器端的酒品信息管理装置，包括：第三获取单元，用于获取酒品信息，所述酒品信息包括酒品图像；第二定位单元，用于根据酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；管理单元，用于基于所述目标位置，对所述酒品信息进行管理。
第六方面，本申请还提供了一种应用终端的酒品信息管理装置，包括：发送单元，用于响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；接收单元，用于接收所述预设服务器发送的所述目标酒品信息；显示单元，用于将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
第七方面，本申请还提供了一种计算机设备，其包括存储器及处理器，所述存储器上存储有计算机程序，所述处理器执行所述计算机程序时执行所述酒品定位方法的步骤，或者执行应用于服务器端的所述酒品信息管理方法的步骤，或者执行应用于终端的所述酒品信息管理方法的步骤。
第八九方面，本申请还提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时使所述处理器实现所述酒品定位方法的步骤，或者实现应用于服务器端的所述酒品信息管理方法的步骤，或者实现应用于终端的所述酒品信息管理方法的步骤。
本申请提供了一种酒品定位方法、酒品信息管理方法、装置、计算机设备 及计算机可读存储介质。本申请通过基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像，基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置，采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。由于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，在复杂的酒窖环境中，能够准确、快速的识别出酒品图像中所包含的酒标，从而能够根据识别出的酒标，再结合酒窖中的预设摄像头的布局，能够实现对酒品的准确定位，以对酒窖中的酒品进行精确的自动化管理，避免由于酒标识别出现差错而导致对酒窖中的酒品进行自动化管理时出现的混乱和错误的情形，提高酒窖中酒品自动化管理的效率和质量，例如，在储存红酒的酒窖中，通过对红酒酒品的准确定位，能够对红酒进行实时、自动及高品质的管理，从而提高了对红酒的自动化管理效率。
附图说明
为了更清楚地说明本申请实施例技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的酒品定位方法的一个应用环境图；
图2为本申请实施例提供的酒品定位方法的一个流程示意图；
图3为本申请实施例提供的酒品定位方法中第一个子流程示意图；
图4为本申请实施例提供的酒品定位方法的第二个子流程示意图；
图5为本申请实施例提供的应用于服务器端的酒品信息管理方法的一个流程示意图；
图6为本申请实施例提供的应用于终端的酒品信息管理方法的一个流程示意图；
图7为本申请实施例提供的酒品定位装置的一个示意性框图；
图8为本申请实施例提供的应用于服务器端的酒品信息管理装置的一个示意性框图；
图9为本申请实施例提供的应用于终端的酒品信息管理装置的一个示意性框图；以及
图10为本申请实施例提供的计算机设备的示意性框图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳 动前提下所获得的所有其他实施例，都属于本申请保护的范围。
应当理解，当在本说明书和所附权利要求书中使用时，术语“包括”和“包含”指示所描述特征、整体、步骤、操作、元素和/或组件的存在，但并不排除一个或多个其它特征、整体、步骤、操作、元素、组件和/或其集合的存在或添加。
请参阅图1，图1为本申请实施例提供的酒品定位方法的一个应用环境图。如图1所示，该应用环境包括：(1)酒窖。酒窖中的不同位置设置有若干个摄像头，如图1所示中，该示例的酒窖中设置有摄像头1、摄像头2、摄像头3及摄像头4，以拍摄酒窖中的酒品，请继续参阅图，如图1所示，在该示例中，包含位置1、位置2及位置3，摄像头1和摄像头2用于拍摄位置1，摄像头3用于拍摄位置2，摄像头4用于拍摄位置3，红酒1酒品可以为存储位置1、位置2或者位置3等酒柜中的酒品，同时，红酒1酒品也可以为进出酒窖出入口的酒品，例如图1中所示，红酒1进入酒窖后，经过位置1、位置2到达位置3，从而放置在位置3，在识别出红酒1所对应的酒标的基础上，结合红酒1所在的位置3以对红酒1所对应的酒品信息进行自动化管理。(2)服务器。用于接收摄像头拍摄并上传的红酒1酒品所对应的酒品图像，对酒品图像进行酒标识别，并根据对酒标进行识所对应的酒标识别结果，再结合摄像头1、摄像头2、摄像头3及摄像头4各自所对应的预设采集位置，尤其最后红酒停留的摄像头4所对应的位置3，对红酒1所对应的酒品进行酒品信息的管理。(3)终端。用于接收用户对红酒1所对应的酒品信息进行查询、编辑、删除等操作，并响应于用户的操作，从服务器获取用户操作所对应的红酒1所对应的酒品信息，可以通过终端上安装的酒品信息管理的APP进行对酒品信息进行操作。
以上各个主体的运行过程如下：1)酒窖中的摄像头拍摄酒窖中的酒品所对应的酒品图像，并将酒品图像上传至服务器，例如，图1中摄像头1和摄像头2拍摄红酒1在位置1所对应的红酒酒品图像1，摄像头3拍摄红酒1在位置2所对应的红酒酒品图像2，摄像头4拍摄红酒1在位置3所对应的红酒酒品图像3，并将红酒酒品图像1、红酒酒品图像2及红酒酒品图像3上传至服务器。2)服务器接收酒品信息，所述酒品信息包括酒品图像，例如接收红酒酒品图像1、红酒酒品图像2或者红酒酒品图像3，获取酒品图像后，基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标，再获取所述摄像头所对应的预设采集位置，可以通过摄像头与位置之间的预设对应关系获取所述摄像头拍摄的酒品所对应的预设采集位置，例如，通过摄像头1和摄像头2进行拍摄所对应的就是位置1，通过摄像头3进行拍摄所对应的就是位置2，也可以为对摄像头拍摄的位置图像进行识别，以识别出拍摄的酒品所对应的预设采集位置，例如，接收到摄像头2拍摄的包含位置1的图像后，对该图像进行识别以识别出该图像对应的位置为位置1，可以是对位置1所对应的位置编号进行识别，并将所述预设采集位置作为 所述目标酒品所对应的当前位置。采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位，对所述酒品图像所对应酒品的酒品信息进行记录、删除、编辑等酒品信息的管理。3)响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置，并将所述目标酒品信息返回至所述终端，接收所述预设服务器发送的所述目标酒品信息，将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
请参阅图2，图2为本申请实施例提供的酒品定位方法的一个流程示意图。如图2所示，该方法包括以下步骤S21-S24：
S21、基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像。
具体地，在酒窖中对酒品进行自动化管理时，会涉及到对酒品进行定位，以得到酒品的位置，根据酒品的位置对酒品所对应的酒品信息进行自动化管理。可以根据对酒品拍摄的实际需求，在酒窖中不同的位置设置摄像头，以通过摄像头拍摄酒窖中的酒品，例如，可以对酒柜中的存酒或者取酒进行拍摄，也可以对酒窖中的存储酒品与取出酒品进行拍摄，摄像头将拍摄的酒品所对应的酒品图像上传至进行酒品定位处理的计算机设备，以使计算机设备获取所述预设摄像头采集的目标酒品所对应的酒品图像。其中，摄像头拍摄的酒品图像，可以为图片，也可以为视频，若摄像头拍摄的为视频，需要从视频中提取视频帧，以得到所述酒品所对应的酒品图像。
S22、基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标。
具体地，需要对酒窖中的酒品进行自动化定位时，需要识别出酒品所属的酒标，即该酒品是什么酒，然后结合该酒品在酒窖中的位置对酒品进行自动化定位。获取到目标酒品所对应的酒品图像后，对所述酒品图像进行图像识别，以识别出所述酒品图像所包含的酒标，从而得到所述酒品图像所对应的酒标。在本申请实施例中，对所述酒品图像进行酒标识别时，基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标，由于OCR文字识别能够快速、准确的识别图像中所包含的规范字符的文字，而深度学习识别又能够准确的识别图像，从而基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，能够准确、快速的识别出酒品图像中所包含的酒标，提高对酒品图像中所包含酒标识别的准确性和识别效率。
S23、获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为 所述目标酒品所对应的当前位置。
具体地，对酒窖中的酒品进行自动化定位时，需要识别出酒品所属的酒标，即该酒品是什么酒，然后结合该酒品在酒窖中的位置对酒品进行自动化定位所得到的该酒品所对应的位置，即酒品位置，对该酒品进行自动化管理，而酒品在酒窖中的位置，即所述摄像头所对应的预设采集位置可以通过以下两种方式实现：
1)可以通过酒窖中的摄像头所对应的固定的预设采集位置而获取，即对每一摄像头预先固定预设采集位置，通过摄像头与固定的预设采集位置之间的对应关系，根据摄像头获取该摄像头所对应的预设采集位置，通过该摄像头即可确定该摄像头所对应的预设采集位置，从而得到所述摄像头所对应的预设采集位置。例如，在酒窖中的A位置设置了摄像头A，通过摄像头A拍摄酒窖中B位置的图像，若接收到摄像头A拍摄的酒品图像，即可知道该酒品图像所对应的酒品在酒窖中的B位置。因此，获取所述预设摄像头采集的目标酒品所对应的酒品图像后，根据上传该酒品图像的摄像头所对应的预设采集位置，即可得到该酒品图像所对应的酒品在拍摄该酒品图像时所在的酒窖中的位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置。
2)可以通过图像检测的方式获取所述摄像头所对应的预设采集位置。即在每一酒品放置位置对应设置一摄像头，通过该摄像头拍摄该放置位置，例如拍摄该放置位置所对应的位置标识，所述位置标识可以为位置编号，检测到酒品在该放置位置时，通过摄像头拍摄获取该放置位置所对应的图像，然后对放置位置所对应的图像进行图像检测，以识别出该放置位置。例如，在酒窖中的A位置设置了摄像头A，通过摄像头A拍摄酒窖中放置酒品的B位置的图像，在对酒品进行定位的过程中，若通过摄像头拍摄到了包含B位置的图像，接收到摄像头A拍摄的包含B位置的图像后，对包含B位置的图像进行图像识别以识别出B位置，即可知道该酒品图像所对应的酒品在酒窖中的位置为B位置，从而获取所述摄像头所对应的预设采集位置。
S24、采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
具体地，获取到所述酒品图像所包含的酒标及所述酒品图像所对应的酒品在酒窖中的当前位置后，即可采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。例如，若一目标酒品所对应的酒品图像所包含的酒标为M，该酒品图像所对应的目标酒品在酒窖中的位置为L，对该目标酒品在酒窖中的定位即可为：酒品M，在酒窖中的L位置。由于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，在复杂的酒窖环境中，能够准确、快速的识别出酒品图像中所包含的酒标，从而能够根据识别出的酒标，再结合酒品在酒窖中的位置，实现对酒品的准确定位，以对酒窖中的 酒品进行精确的自动化管理，避免由于酒标识别出现差错而导致对酒窖中的酒品进行自动化管理时出现的混乱和错误的情形，提高酒窖中酒品自动化管理的效率和质量。
本申请通过基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像，基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置，采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。由于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，在复杂的酒窖环境中，能够准确、快速的识别出酒品图像中所包含的酒标，从而能够根据识别出的酒标，再结合酒窖中的预设摄像头的布局，能够实现对酒品的准确定位，以对酒窖中的酒品进行精确的自动化管理，避免由于酒标识别出现差错而导致对酒窖中的酒品进行自动化管理时出现的混乱和错误的情形，提高酒窖中酒品自动化管理的效率和质量，例如，在储存红酒的酒窖中，通过对红酒酒品的准确定位，能够对红酒进行实时、自动及高品质的管理，从而提高了对红酒的自动化管理效率。
请参阅图3，图3为本申请实施例提供的酒品定位方法的第一个子流程示意图。如图3所示，在该实施例中，所述基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标的步骤，包括以下步骤S31-S33：
S31、根据预设OCR识别方式对所述酒品图像进行OCR识别，以得到所述酒品图像中所包含的文字。
其中，OCR，英文为Optical Character Recognition，光学字符识别，为采用光学的方式将文档中的文字转换成为黑白点阵的图像文件，并通过识别软件将图像中的文字转换成文本格式。
具体地，要进行酒标识别时，通过酒窖中设置的摄像头采集酒窖中的酒品图像，并将酒品图像上传至进行酒标识别的计算机设备，以使所述计算机设备获取待识别的酒品图像，然后所述计算机设备根据预设OCR识别方式对所述酒品图像进行OCR识别，以提取出所述酒品图像中所包含的文字，并将所述文字作为目标文字。由于OCR技术擅长识别规范的文字等字符型内容，因此，可以准确的识别出酒品图像中所包含的文字。例如，在红酒酒窖中对红酒进行酒标自动化识别，以对红酒酒品信息进行管理时，在红酒酒窖中，可以通过红酒酒窖中部署的若干个摄像头拍摄红酒酒品的酒品图像，所述酒品图像可以为通过摄像头拍摄的单张图像(即图片格式的画面)，并将所述酒品图像上传至后台红酒酒品信息服务器，也可以为通过摄像头采集的视频，并将摄像头采集的视频上传至后台红酒酒品信息服务器，然后通过提取所述视频中的视频帧所得到的 图像，后台红酒酒品信息服务器根据预设OCR识别方式对所述酒品图像进行OCR识别，以得到所述酒品图像中所包含的文字，将所述文字作为酒标所对应的目标文字。
进一步地，所述获取酒品图像的步骤之前，还包括：
获取所述酒品图像所对应的初始酒品图像；
对所述初始酒品图像进行分割，以得到所述初始酒品图像中包含预设酒标的局部图像所对应的酒品图像。
具体地，由于不同的物体之间，或者同一物体的不同部分之间，例如，酒品与周围环境之间，不同的酒品之间，或者同一酒品的不同部分之间，会呈现不同的亮度、纹理或者颜色等图像特征的概率是很高的，鉴于此，由于需要管理的酒品是在预设酒品范围之内的，而使需要识别的酒标也是在预设酒标范围之内的，对酒标进行自动识别的步骤之前，可以在获取到所述酒品图像所对应的初始酒品图像之后，在进行酒标自动之前，对获取到的初始酒品图像进行分割，以将初始酒品图像中所包含的酒品之外的环境图像、酒标之外的酒瓶等酒品中其它部分图像进行智能分割，将分割后的包含预设酒标的局部图像当作目标酒标图像的大致位置，将所述目标酒标图像作为进行酒标识别的酒品图像。由于将进行识别的图像聚焦于包含酒标的局部图像，尤其是在酒窖等包含众多环境因素的复杂环境下，能够对酒瓶所对应的酒品图像进行准确的识别，从而得到对应的精准的图像特征，从而能够提高后续进行酒标识别的准确性，并且相比于对整个初始酒品图像进行识别，也减小了数据处理量，从而提高了酒标识别的效率。
S32、根据预设深度学习识别方式对所述酒品图像进行深度学习识别，以提取出所述酒品图像中所包含的图像特征。
其中，深度学习，英文为Deep Learning，在于建立、模拟人脑进行分析学习的神经网络，它模仿人脑的机制来解释数据，深度学习的模型包括Feed forward neural networks(FF or FFNN)and perceptrons(P)、Boltzmann machines(BM)等，可以通过测试样本从众多深度学习模型中筛选出对酒标进行识别效果比较好的深度学习模型。
具体地，获取到酒品图像后，根据预设深度学习识别方式对所述酒品图像进行深度学习识别，由于深度学习为特征学习(或者称为表征学习)，是一种特征工程，从而通过深度学习可以得到所述酒品图像中所包含的图像特征。由于深度学习擅长通过对训练样本图像及所述训练样本图像所对应标签的学习，从而预训练后的基于深度学习的深度学习模型擅长识别图像特征，从而实现可以准确的识别出所述酒品图像中所包含的图像特征，并将所述图像特征作为所述酒品图像所对应的目标图像特征。例如，在对红酒进行红酒酒标自动化识别以对红酒酒品信息管理时，获取到红酒酒品所对应的红酒酒品图像后，根据预设 深度学习识别方式对所述红酒酒品图像进行深度学习识别，以识别出所述红酒酒品图像中所包含的图像特征，并将所述图像特征作为所述红酒酒品图像所对应的目标图像特征。
其中，对所述预设深度学习识别方式所对应的预设深度学习模型进行训练的过程包括以下步骤：
1)对预设酒品所对应的酒标图像进行打标签，以得到酒标图像预训练样本，所述酒标图像预训练样本包括所述训练酒标图像及所述训练酒标图像所对应的训练酒标标签。具体地，是对需要进行管理的酒品所对应的酒标拍摄酒标图像，尤其是从各个角度拍摄该酒品上的酒标图像，并给所拍摄的酒标图像打标签，所谓打标签，即将拍摄的酒标图像和该酒标所对应的酒标名称对应关联起来，以描述该酒标图像对应于该酒标名称，该酒标名称即为该酒标图像所对应的酒标标签。
2)使用所述酒标图像预训练样本对所述预设深度学习模型进行训练。具体地，将酒标图像预训练样本输入至预设深度学习模型，以使预设深度学习模型根据所述训练酒标图像及所述训练酒标图像所对应的训练酒标标签进行自动学习，学习训练酒标图像所对应的训练图像特征和训练酒标标签之间的关联关系，从而完成训练深度学习模型。
3)使用酒标图像测试样本对所述预设深度学习模型进行测试，所述酒标图像测试样本包括所述测试酒标图像及所述测试酒标图像所对应的测试酒标标签。具体地，将训练后的预设深度学习模型采用酒标图像测试样本进行测试，预设深度学习模型对所述测试酒标图像进行识别，以得到测试目标酒标，将所述测试目标酒标与所述测试酒标图像所对应的测试酒标标签进行比较，判断所述测试目标酒标与所述测试酒标图像所对应的测试酒标标签是否一致，若所述测试目标酒标与所述测试酒标图像所对应的测试酒标标签一致，表明深度学习模型对预设酒标识别具有较为准确的识别效果，若所述测试目标酒标与所述测试酒标图像所对应的测试酒标标签不一致，表明深度学习模型对预设酒标识别不具有较为准确的识别效果，以判断预设深度学习模型是否达到预期目标，若达到预期目标，将所述预设深度学习模型投入生产环境，若未达到预期目标，用更多的训练样本、调整训练样本及调整预设深度学习模型重新进行训练，直至预设深度学习模型最终符合预期目标。
S33、根据所述文字与所述图像特征，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标。
具体地，在对酒标进行识别时，由于酒标的范围是可以预先确定的，例如，酒窖中存在哪些品类的酒品是可以预先确定的，因此，可以根据所需要管理的酒品所对应的酒标组成预设酒标数据库，从酒品图像中提取出酒品图像所包含 的文字和图像特征后，可以根据所述文字与所述图像特征，将所述文字与所述图像特征结合起来，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的酒标作为目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标，从而识别出所述酒品图像所对应的酒标，并根据识别出的酒标对酒品信息进行自动化管理。例如，在通过红酒酒窖对酒品信息进行管理时，获取到红酒酒品图像后，识别出所述红酒酒品图像中的文字和图像特征，并根据所述文字和所述图像特征识别出所述红酒酒品图像所对应的红酒酒标，根据识别出的红酒酒标对红酒酒品信息进行管理。
在本申请实施例中，由于酒标一般由文字和图像组成，且不同酒标存在巨大差别，尤其针对红酒酒标，对于不同厂家生产的红酒，其酒标的字体种类、字符大小及语言等均存在巨大的字体差异，尤其在复杂的酒窖环境条件下，且存在多国语言等众多酒标差异的情形，在实现对酒品进行统一自动化管理时，传统技术中仅基于光学文字识别对酒标进行自动识别时，除了字体差异之外，同时还存在对多国文字兼容性较差的问题，会出现对酒标文字识别准确率较低导致识别效果不稳定，而传统的深度学习技术又对训练样本数据量有过高的需求，才能训练出较为准确的深度学习识别模型，基于深度学习技术的图像识别计算量较大，在训练样本不足够大的情形下，也会对酒标识别效率较低，而一般情况下，针对酒品信息管理的应用环境，很难积累起足够大的酒标训练数据样本，因此，导致计算机设备对多种酒标的准确识别异常困难。本申请实施例的申请人在面对复杂环境下对多种酒标存在巨大差异而需要统一进行准确识别的情形，基于酒标由文字和图像组成的特点，将OCR文字识别和深度学习图像识别结合起来实现对酒标进行基于机器视觉的自动识别，由于充分融合了OCR光学识别与深度学习识别的优势，既能够充分利用OCR文字识别对规范字符识别较快且精准的优点，以快速识别所述酒品图像中所包含的字体规范的字符，又能够充分利用深度学习对图像识别较准确的优点，借助于机器视觉的深度学习识别，采用深度学习图像识别的优点克服OCR文字识别的缺点，同时实现对多国语言文字的良好兼容性，从而充分利用OCR文字识别的优点缩小深度学习的搜索范围，以克服深度学习对图像识别的训练数据量要求过高和计算量较大的缺点，同时，由于基于深度学习的图像识别，只需要从较小的搜索范围内进行搜索，可以优化深度学习模型的网络结构，降低系统对计算资源的需求，从而提出通过将光学文字识别与深度学习识别结合起来，综合两种识别方式各自的识别结果，实现对多种酒标，从粗糙到精细、准确的智能识别过程，能够提高对酒标的识别速度和精度，尤其对于红酒酒标存在多语言巨大差异、且酒标训练样本数据量无法足够大，以满足深度学习进行训练的酒标识别情形，可以实现实时、高效地对酒品酒标进行准确的识别，提高了对酒品酒标识别的准确性和效率。
在一实施例中，所述根据所述文字与所述图像特征，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标的步骤包括：
根据所述文字，从预设酒标数据库中筛选出与所述文字相匹配的预设酒标，并将所述预设酒标组成集合，以得到目标酒标集合；
根据所述图像特征，从所述目标酒标集合中筛选出与所述图像特征的匹配度最高的酒标作为目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标。
具体地，从图像中通过OCR识别提取出所述图像中包含的文字后，根据所述文字，从预设酒标数据库中筛选出与所述文字相匹配的预设酒标，并将所述预设酒标组成集合，以得到目标酒标集合，从而充分利用OCR对规范文字识别较快且精准的优点，缩小深度学习搜索酒标的范围至目标酒标集合。
进一步地，一般情况下，由于酒标是文字和图像的结合，或者识别出的文字是不精准的，例如，识别出的文字存在相似、相同或者被包含于其它文字组合中，无法仅通过识别出的文字准确确定酒品图像所对应的酒标，可以在得到所述目标酒标集合后，再根据所述图像特征，从所述目标酒标集合中筛选出与所述图像特征最相匹配的酒标作为目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标，从而识别出所述酒品图像所对应的酒标。由于经过预训练的深度学习模型对图像识别较准确的特点，在深度学习模型搜索酒标范围已经缩小至目标酒标集合中时，相比从预设酒标数据库中直接进行搜索以得到目标酒标，可以大大减小深度学习模型的计算量，并且克服了OCR模型对文字进行识别时，在酒窖等复杂环境中识别准确率较低的缺陷，从而从目标酒标集合中筛选出准确的目标酒标，能够得到准确的所述酒品图像所对应的酒标，提高了识别所述酒品图像所对应酒标的准确性和效率。
请参阅图4，图4为本申请实施例提供的酒品定位方法的第二个子流程示意图。如图4所示，在该实施例中，所述预设摄像头包括若干个摄像头，每个摄像头拍摄不同位置的酒品所对应的酒品图像，或者每个摄像头从不同角度拍摄同一位置的酒品所对应的酒品图像，由于在酒窖中移动酒品行为的多样性，在酒品移动过程中，不可能正好将酒品上需要拍摄的角度正对摄像头以让摄像头拍摄准确、清晰的酒品图像，通过同一位置从不同角度设置摄像头对该位置的酒品进行拍摄，可以避免由于摄像头拍摄角度的问题导致对酒品拍摄不准确或者不清晰，尽可能提高对该位置的酒品拍摄清晰、准确的酒品图像，进而对酒品图像识别准确，从而得到准确度酒品所对应的酒标及酒品位置等酒品信息。当酒品经过该位置或者当酒品放置在该位置时，即可拍摄到该酒品所对应的酒品图像。由于在酒窖中是多个摄像头对酒品进行拍摄，所述酒品图像包括若干张图像，所述获取所述摄像头所对应的预设采集位置，并将所述预设采集位置 作为所述目标酒品所对应的当前位置的步骤包括：
S41、获取所有所述酒品图像各自所对应的预设采集位置，以得到所有所述预设采集位置；
S42、将所有所述预设采集位置按照各自所对应的所述酒品图像的拍摄先后顺序进行排序，以得到采集位置排序队列；
S43、筛选出所述采集位置排序队列中处于末位的末位预设采集位置，并将所述末位预设采集位置作为所述目标酒品所对应的当前位置。
具体地，所述酒品图像包括若干张图像，若干张所述图像通过预设酒窖中预先安装的若干个摄像头获取，若干个所述摄像头可以安装在所述酒窖中的不同预设位置，酒品在酒窖中进行移动时，可以从不同位置拍摄目标酒品，或者从同一位置的不同角度拍摄同一目标酒品。在酒品信息自动管理系统中，一般会在酒窖中设置多个摄像头，例如红酒酒窖自动管理系统中，为了实现对酒窖中红酒酒品的酒品信息自动管理，一般会在酒窖中安装多个摄像头，若干个所述摄像头安装在所述酒窖中的不同预设位置，以从不同位置或者同一位置的不同角度拍摄目标红酒酒品，通过酒窖中预先安装的若干个摄像头获取酒品图片或者视频，从而得到酒品的若干张酒品图像，所述酒品图像包括所述酒品图像的拍摄时间及上传所述酒品图像的摄像头所对应的预设摄像头标识，所述预设摄像头预先设置在预设固定位置，以拍摄所对应的预设采集位置的酒品图像。
得到酒品的若干张图像后，对拍摄得到的酒品图像进行识别，以识别目标酒品的酒标，再获取所述酒品图像所对应的预设采集位置(即可得到酒品当时所处的位置)，根据识别出的酒标、得到的所述酒品图像所对应的预设采集位置及所述酒品图像的拍摄先后顺序(即拍摄时间先后顺序)，将所有所述预设采集位置按照各自所对应的所述酒品图像的拍摄先后顺序进行排序，以得到预设采集位置排序队列，以得到所述酒品所对应的移动轨迹，迭代上述过程，直至所述酒品的位置不再变动，排在最后的位置确定为酒品的最终位置，从而筛选出所述采集位置排序队列中处于末位的末位预设采集位置，并将所述末位预设采集位置作为所述目标酒品所对应的当前位置，实现对酒品进行定位。基于所述酒品的定位，然后对酒品信息进行处理，从而实现对目标酒品的自动化管理。
例如，在红酒酒窖中，对红酒酒品进行定位的过程中，请参阅表格1，表格1为对所有所述酒品图像进行排序后的一个示例，若有ABCDE五个摄像头，A摄像头排在1号位置，B摄像头排在2号位置，C摄像头排在3号位置，D摄像头排在4号位置，E摄像头排在5号位置，将不同的位置连接起来即可形成一条对应的路径，该路径对应于物的移动即为移动轨迹，若酒品依次经过12345位置，即可得到酒品的移动轨迹，若酒品停留在位置5后不再移动，即可以确定酒品移动的终点为位置5，例如，12345为一条路径，134为一条路径，145可以为一条路径，具体的路径可以根据酒窖的布局进行设置。进一步地，若酒窖 中的移动路径具有唯一性，那么根据已知的部分路径可以补全完整的路径，例如，若从1号位置到4号位置只有1234这唯一的路径，即使从已识别的移动轨迹中识别出酒品所对应的移动轨迹为134或者14，也可以根据134或者14确定红酒完整的移动轨迹应该是1234。请继续参阅表格1，表格1如下所示：
表格1
Figure PCTCN2020123637-appb-000001
其中，P ATA1用于描述A摄像头在TA1时刻的酒品图像，P ATA2用于描述A摄像头在TA2时刻的酒品图像，依次类推，可知，A摄像头包含3张识别出包含同一酒标的酒品图像，B摄像头由于被遮挡等情况，导致B摄像头没有包含酒标的图像，C摄像头包含2张识别出包含同一酒标的酒品图像，D摄像头包含1张包含同一酒标的酒品图像，E摄像头包含五张包含同一酒标的酒品图像，上述酒品图像为同一酒品，其中，A摄像头对应1号位置，B摄像头对应2号位置，C摄像头对应3号位置，D摄像头对应4号位置，E摄像头对应5号位置。可知，该酒品的移动轨迹为1345，或者，若路径是唯一的，虽然B摄像头没有包含酒品的图像，但是由于从1至345位置，必须经过B摄像头，可知该酒品的移动轨迹为12345，由于最后识别出酒标的位置在E摄像头所对应的5号位置，且超过了预设时间，其它摄像头均没有识别出该酒标，可知该酒标所对应的酒品最后的位置确定在了E摄像头所对应的5号位置，从而实现根据识别所述红酒酒标所经过的摄像头，以对红酒酒品的行动路线进行追踪，直至确定红酒酒品的最终位置，若该最终位置对应于酒柜的唯一位置，可以更明确的判断所述红酒的位置所对应的酒柜位置，进而实现对红酒的定位。
需要说明的是，上述各个实施例所述的酒标识别方法，可以根据需要将不同实施例中包含的技术特征重新进行组合，以获取组合后的实施方案，但都在本申请要求的保护范围之内。
请参阅图5，图5为本申请实施例提供的应用于服务器端的酒品信息管理方法的一个流程示意图。如图5所示，该方法包括以下步骤S51-S53：
S51、获取酒品信息，所述酒品信息包括酒品图像；
S52、根据以上各个实施例所描述的所述酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；
S53基于所述目标位置，对所述酒品信息进行管理。
具体地，服务器对酒品所对应的酒品信息进行自动化管理时，不但要获取酒品图像，以根据上述各个实施例所描述的酒品定位方法对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的酒品位置，而且要获取所述酒品所对应的酒品年份、酒品产地、生产日期及进出酒窖的进出时间等酒品的其它方面的酒品信息，并在识别出所述酒品图像所对应的酒标及酒品所对应的酒品位置后，基于所述酒标及所述酒品位置，对所述酒品信息进行添加、修改、编辑等记录操作，后续可以查询该酒品在酒窖的什么位置存储着什么样的酒品等，以实现对所述酒品的自动化管理，由于提高了对酒品定位的准确性和效率，基于对酒品定位的准确及快速的识别，对酒窖中的酒品进行自动化管理时，也能够提高对酒品进行自动化管理的效率和准确性，例如，在储存红酒的酒窖中，由于对红酒定位的准确及快速，能够对红酒进行实时、自动及高品质的管理，从而提高了对红酒的自动化管理效率。
请参阅图6，图6为本申请实施例提供的应用于终端的酒品信息管理方法的一个流程示意图。如图6所示，该方法包括以下步骤S61-S63：
S61、响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据以上各个实施例所描述的所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；
S62、接收所述预设服务器发送的所述目标酒品信息；
S63、将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
具体地，用户通过终端对酒品信息进行查询、修改、添加或者删除等编辑操作时，终端响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，，所述酒品位置是根据以上各个实施例所描述的所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置，并将所述目标酒品信息返回至所述终端，终端接收所述预设服务器发送的所述目标酒品信息，将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息，由于提高了对酒品定位识别的准确性和效率，基于对酒品定位的准确及快速的识别，对酒窖中的酒品实现自动化管理时，也能够方便用户实现提高对酒品进行自动化管理的效率、准确性及便捷性。例如，在储存红酒的酒窖中，由于对红酒定位的准确性及快速，能够对红酒进行实时、自动及高品质的管理，，相对于传统技术中通过人工方式对酒窖中的酒品进行查找等酒品管理方式，提高了对红酒的自动化管理效率、便捷性和管理质量。
请参阅图7，图7为本申请实施例提供的酒标定位装置的一个示意性框图。对应于上述所述酒标定位方法，本申请实施例还提供一种酒标定位装置。如图7所示，该酒标定位装置包括用于执行上述所述酒标定位方法的单元，该酒标定位装置可以被配置于服务器等计算机设备中。具体地，请参阅图7，该酒标定位装置70包括第一获取单元71、识别单元72、第二获取单元73及第一定位单元74。
其中，第一获取单元71，用于基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；
识别单元72，用于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；
第二获取单元73，用于获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；
第一定位单元74，用于采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
在一实施例中，所述识别单元72包括：
第一识别子单元，用于根据预设OCR识别方式对所述酒品图像进行OCR识别，以得到所述酒品图像中所包含的文字；
第二识别子单元，用于根据预设深度学习识别方式对所述酒品图像进行深度学习识别，以得到所述酒品图像中所包含的图像特征；
第一筛选子单元，用于根据所述文字与所述图像特征，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标。
在一实施例中，所述第二获取单元73包括：
第一获取子单元，用于获取所有所述酒品图像各自所对应的预设采集位置，以得到所有所述预设采集位置；
排序子单元，用于将所有所述预设采集位置按照各自所对应的所述酒品图像的拍摄先后顺序进行排序，以得到采集位置排序队列；
第二筛选子单元，用于筛选出所述采集位置排序队列中处于末位的末位预设采集位置，并将所述末位预设采集位置作为所述目标酒品所对应的当前位置。
请参阅图8，图8为本申请实施例提供的应用于服务器端的酒品信息管理装置的一个示意性框图。对应于上述应用于服务器端的所述酒品信息管理方法，本申请实施例还提供一种应用于服务器端的酒品信息管理装置。如图8所示，该应用于服务器端的酒品信息管理装置80包括用于执行上述所述应用于服务器端的酒品信息管理方法的单元，该酒品信息管理装置80可以被配置于服务器等计算机设备中。具体地，请参阅图8，该酒品信息管理装置80包括第三获取单元81、第二定位单元82及管理单元83。
其中，第三获取单元81，用于获取酒品信息，所述酒品信息包括酒品图像；第二定位单元82，用于根据以上各个实施例中所描述的所述酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；管理单元83，用于基于所述目标位置，对所述酒品信息进行管理。
请参阅图9，图9为本申请实施例提供的应用于终端的酒品信息管理装置的一个示意性框图。对应于上述应用于终端的所述酒品信息管理方法，本申请实施例还提供一种应用于终端的酒品信息管理装置。如图9所示，该应用于终端的酒品信息管理装置90包括用于执行上述所述应用于终端的酒品信息管理方法的单元，该酒品信息管理装置90可以被配置于智能手机等终端设备中。具体地，请参阅图9，该酒品信息管理装置90包括发送单元91、接收单元92及显示单元93。
其中，发送单元91，用于响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据以上各个实施例所描述的所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；接收单元92，用于接收所述预设服务器发送的所述目标酒品信息；显示单元93，用于将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
本申请实施例还提供一种酒品信息管理系统，所述系统包括酒窖、终端与服务器；其中，所述酒窖中设置有酒柜，酒柜用于放置酒品，酒窖中预设位置设置有摄像头，以通过所述摄像头拍摄所述酒品所对应的酒品图像，并将所述酒品图像上传至所述服务器；所述服务器用于执行以上各个实施例中所描述的应用于服务器端的所述酒品信息管理方法的步骤；所述终端用于执行以上各个实施例中所描述的应用于终端的所述酒品信息管理方法的步骤。
需要说明的是，所属领域的技术人员可以清楚地了解到，上述酒品定位装置和各单元的具体实现过程、上述应用于服务器端的酒品信息管理装置和各单元的具体实现过程、上述应用于终端的酒品信息管理装置和各单元的具体实现过程，可以参考前述各个实施例中各自所对应方法实施例中的相应描述，为了描述的方便和简洁，在此不再赘述。
同时，上述酒品定位装置、上述应用于服务器端的酒品信息管理装置或者上述应用于终端的酒品信息管理装置中各个单元的划分和连接方式仅用于举例说明，在其他实施例中，可将酒品定位装置、上述应用于服务器端的酒品信息管理装置或者上述应用于终端的酒品信息管理装置按照需要划分为不同的单元，也可将酒品定位装置、上述应用于服务器端的酒品信息管理装置或者上述应用于终端的酒品信息管理装置中各单元采取不同的连接顺序和方式，以完成上述 酒品定位装置、上述应用于服务器端的酒品信息管理装置或者上述应用于终端的酒品信息管理装置的全部或部分功能。
上述酒品定位装置、上述应用于服务器端的酒品信息管理装置或者上述应用于终端的酒品信息管理装置可以分别实现为一种计算机程序的形式，该计算机程序可以分别在如图10所示的计算机设备上运行。
请参阅图10，图10是本申请实施例提供的一种计算机设备的示意性框图。该计算机设备500可以是台式机电脑或者服务器等计算机设备，也可以是其他设备中的组件或者部件。
参阅图10，该计算机设备500包括通过系统总线501连接的处理器502、存储器和网络接口505，其中，存储器可以包括非易失性存储介质503和内存储器504，存储器也可以为易失性计算机可读存储介质。
该非易失性存储介质503可存储操作系统5031和计算机程序5032。该计算机程序5032被执行时，可使得处理器502执行一种上述酒品定位方法、上述应用于服务器端的所述酒品信息管理方法或者上述应用于终端的酒品信息管理方法。该处理器502用于提供计算和控制能力，以支撑整个计算机设备500的运行。该内存储器504为非易失性存储介质503中的计算机程序5032的运行提供环境，该计算机程序5032被处理器502执行时，可使得处理器502执行一种上述酒标识别方法、上述应用于服务器端的所述酒品信息管理方法或者上述应用于终端的酒品信息管理方法。该网络接口505用于与其它设备进行网络通信。本领域技术人员可以理解，图10中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备500的限定，具体的计算机设备500可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。例如，在一些实施例中，计算机设备可以仅包括存储器及处理器，在这样的实施例中，存储器及处理器的结构及功能与图10所示实施例一致，在此不再赘述。
在一实施例中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现上述酒品定位方法时，所述处理器502执行如下步骤：基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
在一实施例中，所述处理器502在实现所述基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标的步骤时，具体实现以下步骤：
根据预设OCR识别方式对所述酒品图像进行OCR识别，以得到所述酒品 图像中所包含的文字；
根据预设深度学习识别方式对所述酒品图像进行深度学习识别，以得到所述酒品图像中所包含的图像特征；
根据所述文字与所述图像特征，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标。
在一实施例中，所述预设摄像头包括若干个摄像头，所述酒品图像包括若干张图像，所述处理器502在实现所述获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置的步骤时，具体实现以下步骤：
获取所有所述酒品图像各自所对应的预设采集位置，以得到所有所述预设采集位置；
将所有所述预设采集位置按照各自所对应的所述酒品图像的拍摄先后顺序进行排序，以得到采集位置排序队列；
筛选出所述采集位置排序队列中处于末位的末位预设采集位置，并将所述末位预设采集位置作为所述目标酒品所对应的当前位置。
在一实施例中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现上述应用于服务器端的所述酒品信息管理方法时，所述处理器502执行如下步骤：获取酒品信息，所述酒品信息包括酒品图像；根据以上各个实施例所描述的所述酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；基于所述目标位置，对所述酒品信息进行管理。。
在一实施例中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现上述应用于终端的所述酒品信息管理方法时，所述处理器502执行如下步骤：响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据以上各个实施例所描述的所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；接收所述预设服务器发送的所述目标酒品信息；将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。。
应当理解，在本申请实施例中，处理器502可以是中央处理单元(Central Processing Unit，CPU)，该处理器502还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中，通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
本领域普通技术人员可以理解的是实现上述实施例的方法中的全部或部分流程，是可以通过计算机程序来完成，该计算机程序可存储于一计算机可读存储介质。该计算机程序被该计算机系统中的至少一个处理器执行，以实现上述方法的实施例的流程步骤。
因此，本申请还提供一种计算机可读存储介质。该计算机可读存储介质可以为非易失性的计算机可读存储介质，也可以为易失性计算机可读存储介质，该计算机可读存储介质存储有计算机程序，该计算机程序被处理器执行时使处理器执行以上各实施例中所描述的所述酒品定位方法的步骤，或者使处理器执行以上各实施例中所描述的上述应用于服务器端的所述酒品信息管理方法，或者使处理器执行以上各实施例中所描述的上述应用于终端的酒品信息管理方法。
所述计算机可读存储介质可以是前述设备的内部存储单元，例如设备的硬盘或内存。所述计算机可读存储介质也可以是所述设备的外部存储设备，例如所述设备上配备的插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)等。进一步地，所述计算机可读存储介质还可以既包括所述设备的内部存储单元也包括外部存储设备。
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，上述描述的设备、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
所述存储介质为实体的、非瞬时性的存储介质，例如可以是U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、磁碟或者光盘等各种可以存储计算机程序的实体存储介质。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
在本申请所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的。例如，各个单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式。例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。
本申请实施例方法中的步骤可以根据实际需要进行顺序调整、合并和删减。本申请实施例装置中的单元可以根据实际需要进行合并、划分和删减。另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各 个单元单独物理存在，也可以是两个或两个以上单元集成在一个单元中。
该集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分，或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台电子设备(可以是个人计算机，终端，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。
以上所述，仅为本申请的具体实施方式，但本申请明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。

Claims (10)

  1. 一种酒品定位方法，其特征在于，包括：
    基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；
    基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；
    获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；
    采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
  2. 根据权利要求1所述酒品定位方法，其特征在于，所述基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标的步骤包括：
    根据预设OCR识别方式对所述酒品图像进行OCR识别，以得到所述酒品图像中所包含的文字；
    根据预设深度学习识别方式对所述酒品图像进行深度学习识别，以得到所述酒品图像中所包含的图像特征；
    根据所述文字与所述图像特征，从预设酒标数据库中筛选出与所述文字及所述图像特征相匹配的目标酒标，并将所述目标酒标作为所述酒品图像所对应的酒标。
  3. 根据权利要求1或者2所述酒品定位方法，其特征在于，所述预设摄像头包括若干个摄像头，所述酒品图像包括若干张图像，所述获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置的步骤包括：
    获取所有所述酒品图像各自所对应的预设采集位置，以得到所有所述预设采集位置；
    将所有所述预设采集位置按照各自所对应的所述酒品图像的拍摄先后顺序进行排序，以得到采集位置排序队列；
    筛选出所述采集位置排序队列中处于末位的末位预设采集位置，并将所述末位预设采集位置作为所述目标酒品所对应的当前位置。
  4. 一种酒品信息管理方法，应用于服务器端，其特征在于，包括：
    获取酒品信息，所述酒品信息包括酒品图像；
    根据如权利要求1-3任一项所述酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；
    基于所述目标位置，对所述酒品信息进行管理。
  5. 一种酒品信息管理方法，应用于终端，其特征在于，包括：
    响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据如权利 要求1-3任一项所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；
    接收所述预设服务器发送的所述目标酒品信息；
    将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
  6. 一种酒品定位装置，其特征在于，包括：
    第一获取单元，用于基于酒窖中预设摄像头，获取所述预设摄像头采集的目标酒品所对应的酒品图像；
    识别单元，用于基于OCR文字识别与深度学习识别相结合的预设酒标识别方法，对所述酒品图像进行识别，以得到所述酒品图像所对应的酒标；
    第二获取单元，用于获取所述摄像头所对应的预设采集位置，并将所述预设采集位置作为所述目标酒品所对应的当前位置；
    第一定位单元，用于采用所述酒标及所述当前位置描述所述目标酒品所对应的位置，以对所述目标酒品进行定位。
  7. 一种酒品信息管理装置，应用于服务器端，其特征在于，包括：
    第三获取单元，用于获取酒品信息，所述酒品信息包括酒品图像；
    第二定位单元，用于根据如权利要求1-3任一项所述酒品定位方法，对所述酒品图像所对应的酒品进行定位，以得到所述酒品所对应的目标位置；
    管理单元，用于基于所述目标位置，对所述酒品信息进行管理。
  8. 一种酒品信息管理装置，应用于终端，其特征在于，包括：
    发送单元，用于响应于用户操作，发送酒品信息获取请求至预设服务器，以使所述预设服务器根据所述酒品信息获取请求，获取所述酒品信息获取请求所对应的目标酒品信息，并将所述目标酒品信息返回至所述终端，其中，所述目标酒品信息包含基于酒品所对应的酒品位置所关联的酒品信息，所述酒品位置是根据如权利要求1-3任一项所述酒品定位方法对所述酒品进行定位而得到的所述酒品所在的位置；
    接收单元，用于接收所述预设服务器发送的所述目标酒品信息；
    显示单元，用于将所述目标酒品信息进行显示，以使所述用户获取所述目标酒品信息。
  9. 一种计算机设备，其特征在于，所述计算机设备包括存储器以及与所述存储器相连的处理器；所述存储器用于存储计算机程序；所述处理器用于运行所述计算机程序，以执行如权利要求1-3任一项所述方法的步骤，或者执行如权利要求4所述方法的步骤，或者执行如权利要求5所述方法的步骤。
  10. 一种计算机可读存储介质，其特征在于，所述存储介质存储有计算机程序，所述计算机程序被处理器执行时，实现如权利要求1-3中任一项所述方法的步骤，或者实现如权利要求4所述方法的步骤，或者实现如权利要求5所述方法的步骤。
PCT/CN2020/123637 2020-09-28 2020-10-26 酒品定位方法、酒品信息管理方法、装置、设备及存储介质 WO2022062027A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
GB2304282.3A GB2613753A (en) 2020-09-28 2020-10-26 Wine product positioning method and apparatus, wine product information management method and apparatus, and device and storage medium
JP2023543250A JP7502570B2 (ja) 2020-09-28 2020-10-26 酒製品位置決め方法、酒製品情報管理方法及びその装置、デバイス及び記憶媒体
US18/127,006 US20230237825A1 (en) 2020-09-28 2023-03-28 Wine product positioning method, wine product information management method and apparatus, device, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011040743.3 2020-09-28
CN202011040743.3A CN112161984B (zh) 2020-09-28 2020-09-28 酒品定位方法、酒品信息管理方法、装置、设备及存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/127,006 Continuation US20230237825A1 (en) 2020-09-28 2023-03-28 Wine product positioning method, wine product information management method and apparatus, device, and storage medium

Publications (1)

Publication Number Publication Date
WO2022062027A1 true WO2022062027A1 (zh) 2022-03-31

Family

ID=73861412

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/123637 WO2022062027A1 (zh) 2020-09-28 2020-10-26 酒品定位方法、酒品信息管理方法、装置、设备及存储介质

Country Status (5)

Country Link
US (1) US20230237825A1 (zh)
JP (1) JP7502570B2 (zh)
CN (1) CN112161984B (zh)
GB (1) GB2613753A (zh)
WO (1) WO2022062027A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113610542A (zh) * 2021-07-27 2021-11-05 青岛博芬智能科技股份有限公司 用于酒品信息获取的方法及装置、云端服务器和存储介质
CN114120123A (zh) * 2021-11-29 2022-03-01 青岛海信智慧生活科技股份有限公司 酒品信息管理方法、智能酒柜及系统
CN115018433B (zh) * 2022-08-10 2022-11-29 四川港投新通道物流产业投资集团有限公司 一种酒品供应链监控方法、装置、设备及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105588543A (zh) * 2014-10-22 2016-05-18 中兴通讯股份有限公司 一种基于摄像头实现定位的方法、装置及定位系统
US20190294896A1 (en) * 2016-12-19 2019-09-26 Waymo Llc Object detection neural networks
CN110807431A (zh) * 2019-11-06 2020-02-18 上海眼控科技股份有限公司 对象定位方法、装置、电子设备及存储介质
CN111401461A (zh) * 2020-03-24 2020-07-10 郭俊 酒品信息管理方法、装置、计算机设备以及存储介质

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2474177A (en) * 2008-06-24 2011-04-06 Yvonne Sylvia Griffits Computer controlled object locating system
CN105589925A (zh) * 2015-11-25 2016-05-18 小米科技有限责任公司 信息推荐方法、装置及系统
JP6770464B2 (ja) 2017-03-08 2020-10-14 株式会社日立製作所 製造管理方法、及び製造管理システム
SG11202013090PA (en) 2018-07-04 2021-01-28 Mitsubishi Heavy Industries Machinery Systems Ltd Vehicle number identification device, vehicle number identification method, and program
CN109376731A (zh) * 2018-08-24 2019-02-22 北京三快在线科技有限公司 一种文字识别方法和装置
CN110427936B (zh) * 2019-07-04 2022-09-30 深圳市新潮酒窖文化传播有限公司 一种酒窖的藏酒管理方法及系统
CN111027450A (zh) * 2019-12-04 2020-04-17 深圳市新国都金服技术有限公司 银行卡信息识别方法、装置、计算机设备及存储介质
CN111401444B (zh) * 2020-03-16 2023-11-03 深圳海关食品检验检疫技术中心 红酒原产地的预测方法、装置、计算机设备及存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105588543A (zh) * 2014-10-22 2016-05-18 中兴通讯股份有限公司 一种基于摄像头实现定位的方法、装置及定位系统
US20190294896A1 (en) * 2016-12-19 2019-09-26 Waymo Llc Object detection neural networks
CN110807431A (zh) * 2019-11-06 2020-02-18 上海眼控科技股份有限公司 对象定位方法、装置、电子设备及存储介质
CN111401461A (zh) * 2020-03-24 2020-07-10 郭俊 酒品信息管理方法、装置、计算机设备以及存储介质

Also Published As

Publication number Publication date
GB2613753A (en) 2023-06-14
CN112161984A (zh) 2021-01-01
GB2613753A8 (en) 2023-06-28
CN112161984B (zh) 2022-03-08
US20230237825A1 (en) 2023-07-27
JP7502570B2 (ja) 2024-06-18
GB202304282D0 (en) 2023-05-10
JP2023543639A (ja) 2023-10-17

Similar Documents

Publication Publication Date Title
WO2022062027A1 (zh) 酒品定位方法、酒品信息管理方法、装置、设备及存储介质
CN109284729B (zh) 基于视频获取人脸识别模型训练数据的方法、装置和介质
US11270099B2 (en) Method and apparatus for generating facial feature
WO2020253127A1 (zh) 脸部特征提取模型训练方法、脸部特征提取方法、装置、设备及存储介质
EP3772036A1 (en) Detection of near-duplicate image
WO2018170663A1 (zh) 图像标注方法、装置及电子设备
CN112115950B (zh) 酒标识别方法、酒品信息管理方法、装置、设备及存储介质
WO2021042505A1 (zh) 基于文字识别技术的笔记生成方法、装置和计算机设备
WO2022089170A1 (zh) 字幕区域识别方法、装置、设备及存储介质
CN112084812B (zh) 图像处理方法、装置、计算机设备及存储介质
CN113763348A (zh) 图像质量确定方法、装置、电子设备及存储介质
KR102440198B1 (ko) 시각 검색 방법, 장치, 컴퓨터 기기 및 저장 매체 (video search method and apparatus, computer device, and storage medium)
CN115661846A (zh) 数据处理方法、装置、电子设备和存储介质
CN107656760A (zh) 数据处理方法及装置、电子设备
CN110503087A (zh) 一种拍照框题的搜索方法、装置、终端及存储介质
CN115457585A (zh) 作业批改的处理方法、装置、计算机设备及可读存储介质
CN116233534A (zh) 视频处理方法及装置、电子设备和存储介质
CN112560728B (zh) 目标对象识别方法及装置
CN110381367B (zh) 一种视频处理方法、设备及计算机可读存储介质
CN113780116A (zh) 发票分类方法、装置、计算机设备和存储介质
CN113657187A (zh) 一种脸部识别方法、设备和计算机可读存储介质
CN111461248A (zh) 一种摄影构图线匹配方法、装置、设备及存储介质
CN111597906B (zh) 一种结合文字信息的快速绘本识别方法及系统
CN114663868A (zh) 基于组合学习的视频平台logo检测方法及系统及设备
CN117315704A (zh) 字段识别方法、装置、计算机设备及计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20954873

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 202304282

Country of ref document: GB

Kind code of ref document: A

Free format text: PCT FILING DATE = 20201026

ENP Entry into the national phase

Ref document number: 2023543250

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20954873

Country of ref document: EP

Kind code of ref document: A1