Search

CN-115830708-B - Image data processing method, device, medium and product

CN115830708BCN 115830708 BCN115830708 BCN 115830708BCN-115830708-B

Abstract

The embodiment of the disclosure discloses an image data processing method, device, medium and product, wherein the method comprises the steps of acquiring image data to be processed acquired by a showcase, identifying hands in target image areas of each of a plurality of images to be processed, determining a plurality of target images in the plurality of images to be processed according to a hand identification result, acquiring hand movement distances corresponding to any two adjacent target images in the plurality of target images, and performing frame extraction on the plurality of target images according to the hand movement distances to acquire a plurality of frame extraction target images. According to the technical scheme, on the premise of reducing the number of the plurality of frame extraction target images as much as possible, the image data obtained based on frame extraction can be ensured to obtain a relatively coherent hand movement track, the data size of the image data to be processed is reduced, the accuracy of determining articles moved out of or into a showcase is improved, the data processing cost is reduced, and the user experience is improved.

Inventors

  • XIONG WENTAO

Assignees

  • 元气森林(北京)食品科技集团有限公司

Dates

Publication Date
20260505
Application Date
20221117

Claims (9)

  1. 1. A method of processing image data, the method comprising: Acquiring to-be-processed image data acquired by a showcase, wherein the to-be-processed image data comprises a plurality of to-be-processed images arranged according to acquisition time; Identifying a hand in a target image area of each of the plurality of images to be processed, the target image area including an item access of the display case; Determining a plurality of target images in the plurality of images to be processed according to a hand recognition result, and acquiring hand movement distances corresponding to any two adjacent target images in the plurality of target images, wherein the target images are corresponding target image areas comprising the images to be processed of the hands; the method comprises the steps of obtaining hand movement distances corresponding to any two adjacent target images, wherein the hand movement distances comprise determining that the distance of one hand moving in any two adjacent target images is the hand movement distance corresponding to the any two adjacent target images in response to that only one hand is included in each target image area of the any two adjacent target images, or determining that the target hand with the longest movement distance in any two adjacent target images in the plurality of hands is the hand movement distance corresponding to the any two adjacent target images in response to that the plurality of hands are included in each target image area of the any two adjacent target images; And extracting frames from the plurality of target images according to the hand movement distance to obtain a plurality of frame extraction target images, wherein the hand movement distance corresponding to any two adjacent frame extraction target images in the plurality of frame extraction target images belongs to a hand movement distance interval.
  2. 2. The image data processing method according to claim 1, characterized in that the method further comprises: Obtaining the number of frame extraction obtained by subtracting the number of the plurality of frame extraction target images from the number of the plurality of target images; And in response to the frame extraction number being smaller than or equal to a preset frame extraction number, carrying out frame supplementing on the plurality of frame extraction target images so as to obtain a plurality of frame supplementing target images.
  3. 3. The image data processing method according to claim 2, wherein a number of complementary frames obtained by subtracting the number of the plurality of frame-drawing target images from the number of the plurality of frame-drawing target images is greater than or equal to a number difference obtained by subtracting the number of frame-drawing from the preset frame-drawing number.
  4. 4. A method of processing image data according to any one of claims 1 to 3, wherein the method further comprises: Acquiring cabinet door opening indication information, wherein the cabinet door opening indication information is used for indicating that a cabinet door of the showcase is opened; And responding to the cabinet door opening indication information, and acquiring the image data to be processed.
  5. 5. The image data processing method according to claim 4, characterized in that the method further comprises: Determining a plurality of first to-be-extracted frame images in the plurality of to-be-processed images according to the hand recognition result, wherein the first to-be-extracted frame images are to-be-processed images of which the corresponding target image area does not comprise hands and the hands are positioned outside an article storage area of the display cabinet; And extracting frames from the plurality of first frame images to be extracted according to the acquisition time to acquire a plurality of first frame images, wherein the time difference between the acquisition times corresponding to any two adjacent first frame images in the plurality of first frame images belongs to a first acquisition time difference interval.
  6. 6. The image data processing method according to claim 5, characterized in that the method further comprises: Determining a plurality of second to-be-extracted frame images in the plurality of to-be-processed images according to the hand recognition result, wherein the second to-be-extracted frame images are to-be-processed images which do not comprise hands in corresponding target image areas and are positioned in an article storage area of a display cabinet; and extracting frames from the plurality of second frame images to be extracted according to the acquisition time to acquire a plurality of second frame images, wherein the time difference between the acquisition times corresponding to any two adjacent second frame images in the plurality of second frame images belongs to a second acquisition time difference interval.
  7. 7. An electronic device comprising a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to implement the method of any of claims 1-6.
  8. 8. A computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the method of any of claims 1-6.
  9. 9. A computer program product comprising computer instructions which, when executed by a processor, implement the method of any of claims 1-6.

Description

Image data processing method, device, medium and product Technical Field The present disclosure relates to the field of control technologies, and in particular, to an image data processing method, apparatus, medium, and product. Background In recent years, when a merchant or an enterprise stores articles, in order to facilitate a user to know information of the articles, the articles may be placed in a showcase, so as to store and display the articles at the same time. When a user needs to remove an item from the display case or insert an item into the display case, the user can open the display case by himself and perform a corresponding operation. In such a scenario, a merchant or business needs to monitor the items in the display case, determine the users moving out of or into the display case, and the items moved out of or into the display case, and settle. Disclosure of Invention The embodiment of the disclosure provides an image data processing method, device, medium and product. In a first aspect, an embodiment of the present disclosure provides an image data processing method. Specifically, the image data processing method includes: Acquiring to-be-processed image data acquired by a showcase, wherein the to-be-processed image data comprises a plurality of to-be-processed images arranged according to acquisition time; Identifying a hand in a target image area of each of the plurality of images to be processed, the target image area including an item access of the display case; Determining a plurality of target images in a plurality of images to be processed according to a hand recognition result, and acquiring hand movement distances corresponding to any two adjacent target images in the plurality of target images, wherein the target images are the images to be processed of the hands in the corresponding target image areas; And carrying out frame extraction on the plurality of target images according to the hand movement distance to acquire a plurality of frame extraction target images, wherein the hand movement distance corresponding to any two adjacent frame extraction target images in the plurality of frame extraction target images belongs to a hand movement distance interval. In one embodiment of the present disclosure, the method further comprises: obtaining the number of frame extraction obtained by subtracting the number of the plurality of frame extraction target images from the number of the plurality of target images; And in response to the frame extraction number being smaller than or equal to the preset frame extraction number, carrying out frame supplementing on the plurality of frame extraction target images so as to obtain a plurality of frame supplementing target images. In one embodiment of the present disclosure, the number of complementary frames obtained by subtracting the number of the plurality of frame-extracting target images from the number of the plurality of frame-compensating target images is greater than or equal to a difference in number obtained by subtracting the number of frame-extracting from the preset number of frame-extracting. In one embodiment of the present disclosure, acquiring a hand movement distance corresponding to any two adjacent target images includes: In response to that only one hand is included in each target image area of any two adjacent target images, determining the moving distance of the hand in any two adjacent target images as the corresponding hand moving distance of any two adjacent target images; Or, in response to the fact that the target image areas of any two adjacent target images respectively comprise a plurality of hands, determining the target hand with the longest moving distance in any two adjacent target images in the plurality of hands, and determining the moving distance of the target hand in any two adjacent target images as the corresponding hand moving distance of any two adjacent target images. In one embodiment of the present disclosure, the method further comprises: Acquiring cabinet door opening indication information, wherein the cabinet door opening indication information is used for indicating that a cabinet door of a showcase is opened; and responding to the cabinet door opening indication information, and acquiring the image data to be processed. In one embodiment of the present disclosure, the method further comprises: Determining a plurality of first to-be-extracted frame images in the plurality of to-be-processed images according to the hand recognition result, wherein the first to-be-extracted frame images are to-be-processed images of which the corresponding target image area does not comprise hands and the hands are positioned outside the article storage area of the showcase; And carrying out frame extraction on the plurality of first frame images to be extracted according to the acquisition time so as to acquire the plurality of first frame images, wherein the time difference between the acquisition times corr