CN112801681A - Product popularity trend analysis method and device, electronic equipment and storage medium - Google Patents
Product popularity trend analysis method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112801681A CN112801681A CN201911102019.6A CN201911102019A CN112801681A CN 112801681 A CN112801681 A CN 112801681A CN 201911102019 A CN201911102019 A CN 201911102019A CN 112801681 A CN112801681 A CN 112801681A
- Authority
- CN
- China
- Prior art keywords
- image
- pedestrian
- target product
- target
- images
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
-
- 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
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Abstract
The embodiment of the invention provides a method and a device for analyzing a product popularity trend, electronic equipment and a storage medium, wherein the method comprises the following steps: the method comprises the steps of obtaining at least two pedestrian images collected by video monitoring equipment, identifying the attribute value of each visual attribute of a target product in a target product image extracted from each pedestrian image, and determining the popularity trend of the target product based on the attribute value of the visual attribute of the target product in each target product image. Because the pedestrians in the at least two pedestrian images are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and represent the wearing tendency of the pedestrians, the popular trend of the target products is analyzed based on the target products in the pedestrian images, the analysis of the popular trend by using a data set containing the exchange data and/or the return data can be avoided, and the accuracy of the analysis of the popular trend of the products is improved.
Description
Technical Field
The invention relates to the technical field of big data analysis, in particular to a method and a device for analyzing a product popularity trend, electronic equipment and a storage medium.
Background
Popular trend analysis of products, such as the clothing industry, is based on big data technology. In the related art, it is common to acquire purchase information for purchasing clothing from an e-commerce website and then perform a popularity trend analysis based on the purchase information.
However, the inventor finds that, in addition to the purchase data of the purchased clothes, a large amount of exchange data and/or return data are generally available in the purchase information acquired from the e-commerce website, and the existence of the exchange data and/or the return data can cause errors in analysis results when clothes fashion trend analysis is performed, so that the analysis results are not accurate enough.
Disclosure of Invention
An object of the embodiments of the present invention is to provide a method and an apparatus for analyzing a product popularity trend, an electronic device, and a storage medium, so as to avoid using a data set including exchange data and/or return data to perform popularity trend analysis, and improve accuracy of the product popularity trend analysis. The specific technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a method for analyzing a product popularity trend, where the method includes:
acquiring an image set of pedestrian images acquired by video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
extracting a target product image from each pedestrian image;
identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image;
and determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
Optionally, the acquiring an image set of pedestrian images acquired by the video monitoring device includes:
acquiring a video acquired by video monitoring equipment by adopting a multi-target tracking algorithm, wherein each video frame in the video is marked with a position of a target frame and a target label;
and extracting a video frame from the video, and acquiring a pedestrian image containing a pedestrian from the extracted video frame based on the target label and the position of the target frame of the extracted video frame.
Optionally, extracting a video frame from the video, and acquiring a pedestrian image including a pedestrian from the extracted video frame based on the target tag and the position of the target frame of the extracted video frame, includes:
extracting a plurality of video frames from a video;
for each video frame, extracting pedestrian images corresponding to different contained pedestrians from the video frame based on the positions of the target labels and the target frames of the extracted video frame;
acquiring a plurality of pedestrian images with the same pedestrian corresponding to the plurality of video frames, and performing quality analysis on the plurality of pedestrian images to obtain quality scores corresponding to the plurality of pedestrian images respectively;
and selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
Optionally, extracting the target product image from each pedestrian image includes:
segmenting the pedestrian image by adopting a target product segmentation algorithm to obtain target product images corresponding to target products at different wearing parts in the pedestrian image;
optionally, identifying each target product image to obtain an attribute value of each visual attribute of the target product in each target product image, includes:
performing attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of the target product in each target product image;
and analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
Optionally, determining the popularity trend of the target product based on the attribute values of the respective visual attributes of the target product in each target product image, includes:
determining a target product corresponding to the visual attribute with the maximum attribute value as the most popular product based on the attribute values of the visual attributes; and/or
And determining that the target product corresponding to the visual attribute with the attribute value sequence positioned before the preset digit is a popular product based on the sequence of the attribute values of the visual attributes from large to small.
In a second aspect, an embodiment of the present invention further provides an apparatus for analyzing a product popularity trend, where the apparatus includes:
the image set acquisition module is used for acquiring an image set of pedestrian images acquired by the video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
the image extraction module is used for extracting a target product image from each pedestrian image;
the attribute value identification module is used for identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image;
and the popularity trend analysis module is used for determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
Optionally, the image set obtaining module includes:
the video acquisition submodule is used for acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm, wherein each video frame in the video is marked with a position of a target frame and a target label;
and the image set acquisition module submodule is used for extracting the video frames from the video and acquiring pedestrian images containing pedestrians from the extracted video frames on the basis of the target labels and the positions of the target frames of the extracted video frames.
Optionally, the image set obtaining module sub-module includes:
a video frame extraction unit for extracting a plurality of video frames from a video;
an image segmentation unit for extracting, for each video frame, pedestrian images corresponding to different contained pedestrians from the video frame based on the extracted target tags and positions of the target frames of the video frame;
the quality analysis unit is used for acquiring a plurality of pedestrian images which are corresponding to the video frames and have the same pedestrian, and performing quality analysis on the pedestrian images to obtain quality scores which respectively correspond to the pedestrian images;
and the image set acquisition unit is used for selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
Optionally, the image extraction module is specifically configured to:
segmenting the pedestrian image by adopting a target product segmentation algorithm to obtain target product images corresponding to target products at different wearing parts in the pedestrian image;
optionally, the attribute value identification module includes:
the visual attribute identification submodule is used for carrying out attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of the target product in each target product image;
and the attribute value analysis submodule is used for analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
Optionally, the popularity trend analysis module is specifically configured to:
determining a target product corresponding to the visual attribute with the maximum attribute value as the most popular product based on the attribute values of the visual attributes; and/or
And determining that the target product corresponding to the visual attribute with the attribute value sequence positioned before the preset digit is a popular product based on the sequence of the attribute values of the visual attributes from large to small.
According to the method, the device, the electronic equipment and the storage medium for analyzing the product popularity trend, when the popularity trend of the target product is determined, an image set of pedestrian images acquired by monitoring equipment can be acquired firstly, then the target product image is extracted from each pedestrian image, each target product image is identified, the attribute value of each visual attribute of the target product in each target product image is acquired, and finally the popularity trend of the target product is determined based on the attribute value of each visual attribute of the target product in each target product image. Because the pedestrians in at least two pedestrian images in the image set are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images in the image set are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and can represent the wearing tendency of the pedestrians, the popular tendency of the target products is analyzed based on the target products in the pedestrian images, the popular tendency analysis by using a data set containing the goods change data and/or the goods return data can be avoided, and the accuracy of the popular tendency analysis of the products is improved. Of course, not all of the advantages described above need to be achieved at the same time in the practice of any one product or method of the invention.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of a first implementation of a method for analyzing product popularity trend according to an embodiment of the present invention;
FIG. 2 is a flowchart of a second embodiment of a method for analyzing popularity trend of a product according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a third embodiment of a method for analyzing popularity trend of a product according to an embodiment of the present invention;
FIG. 4 is a flowchart of a fourth implementation of a method for analyzing product popularity trend according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an apparatus for analyzing product popularity trend according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the prior art, the purchasing information of purchased clothes is generally obtained from an e-commerce website, and then the fashion trend analysis is performed based on the purchasing information, however, in addition to the purchasing data of purchased clothes, the purchasing information obtained from the e-commerce website generally has a large amount of exchange data and/or return data, and the existence of the exchange data and/or the return data may cause errors in the analysis result when the fashion trend analysis of clothes is performed, so that the analysis result is not accurate enough.
In view of the above, embodiments of the present invention provide a method and an apparatus for analyzing a product popularity trend, an electronic device, and a storage medium, so as to obtain an attribute value of each visual attribute of a target product in each target product image by acquiring a pedestrian image photographed at an entrance/exit of a business center, extracting a target product image from each pedestrian image, identifying each target product image, and finally determining a popularity trend of the target product based on the attribute value of each visual attribute of the target product in each target product image. Therefore, the method can avoid using a data set containing the exchange data and/or the return data to perform the popular trend analysis, and improve the accuracy of the popular trend analysis of the product.
Next, a method for analyzing a product popularity trend according to an embodiment of the present invention is described first, and as shown in fig. 1, is a flowchart of a first implementation manner of the method for analyzing a product popularity trend according to an embodiment of the present invention, where the method may include:
and S110, acquiring an image set of pedestrian images acquired by the video monitoring equipment.
In some examples, the video monitoring device may be installed at an entrance or other busy areas of various malls in a city, and then the video monitoring device may collect images of pedestrians, so that the images of pedestrians collected by the video monitoring device may be acquired, thereby forming an image set including a plurality of images of pedestrians.
In some examples, in the image set, pedestrians in at least two pedestrian images in the image set are different, and the ratio of the number of pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set include the target product image;
for example, the image set may include 3 images of the pedestrian a, 4 images of the pedestrian B, and 5 images of the pedestrian C, where a ratio of the number of images of the pedestrian a to the number of images of the pedestrian B is within a preset ratio range, a ratio of the number of images of the pedestrian a to the number of images of the pedestrian C is within a preset ratio range, and a ratio of the number of images of the pedestrian C to the number of images of the pedestrian B is within a preset ratio range. In still other examples, the smaller the preset ratio range is set, the closer the number of pedestrian images of the respective pedestrians.
In still other examples, the preset ratio range may be a ratio range that is preset empirically. For example, the range of the amount of the organic compound is 0.8 to 1.2, or 0.9 to 1.1.
In some examples, the target product may be a product that analyzes a popularity trend using the method for analyzing a popularity trend of a product according to an embodiment of the present invention, and the target product may include: jackets, pants, jackets, watches, glasses, hats, masks, and the like.
And S120, extracting a target product image from each pedestrian image.
After an image set containing a plurality of pedestrian images is acquired, a target product image may be extracted from each pedestrian image.
In some examples, when the popularity trend of a plurality of products is analyzed, and the target product is a plurality of products, a plurality of target product images can be respectively extracted for each target product, wherein the target product images are respectively in a plurality of pedestrian images.
S130, identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image.
And S140, determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
In some examples, after obtaining the target product image, in order to determine the popularity trend of the target product, each target product image may be identified first, so as to identify the attribute value of the visual attribute of the target product in the target product image.
In some examples, the visual attributes may include: style, color, brand, etc.
In still other examples, each target product image may be input into a pre-trained recognition model for analysis, so that the attribute values of the visual attributes of the target product in each target product image may be obtained. The recognition model obtained by pre-training can be obtained by training by adopting a product image sample labeled with visual attributes.
After the attribute values of the visual attributes of the target product in each target product image are obtained, the popularity trend of the target product can be determined based on the attribute values of the visual attributes of the target product in each target product image.
In some examples, the popularity trend of the target product may refer to the most popular product, which may refer to attributes of the product, e.g., style, color.
In still other examples, the popularity trend of the target product may also refer to some products that are relatively popular.
Therefore, when determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image, the target product corresponding to the visual attribute with the largest attribute value may be determined as the most popular product based on the attribute values of the visual attributes, and/or the target product corresponding to the visual attribute with the attribute value ranking before the preset number of digits may be determined as the popular product based on the ranking of the attribute values of the visual attributes from large to small.
The preset digit may be a digit preset manually, for example, the preset digit may be 3 bits, or may be 10 bits, and the like. When the preset digit is 3 digits, the target product corresponding to the visual attribute with the attribute value sorted to the first 3 digits is determined to be a popular product based on the sorting of the attribute values of the visual attributes from large to small.
According to the method for analyzing the product popularity trend provided by the embodiment of the invention, when the popularity trend of the target product is determined, an image set of pedestrian images acquired by monitoring equipment can be firstly acquired, then the target product image is extracted from each pedestrian image, each target product image is identified, the attribute value of each visual attribute of the target product in each target product image is acquired, and finally, the popularity trend of the target product is determined based on the attribute value of each visual attribute of the target product in each target product image. Because the pedestrians in at least two pedestrian images in the image set are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images in the image set are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and can represent the wearing tendency of the pedestrians, the popular tendency of the target products is analyzed based on the target products in the pedestrian images, the popular tendency analysis by using a data set containing the goods change data and/or the goods return data can be avoided, and the accuracy of the popular tendency analysis of the products is improved.
On the basis of the method for analyzing the product popularity trend shown in fig. 1, an embodiment of the present invention further provides a possible implementation manner, as shown in fig. 2, which is a flowchart of a second implementation manner of the method for analyzing the product popularity trend according to the embodiment of the present invention, where the method may include:
s210, acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm.
And each video frame in the video is marked with the position of the target frame and the target label.
S220, extracting a video frame from the video, and acquiring a pedestrian image containing a pedestrian from the extracted video frame based on the target label and the position of the target frame of the extracted video frame.
In some examples, the video monitoring device may capture images of pedestrians and may also capture video. In order to obtain a plurality of target product images from one pedestrian image, the video monitoring device can adopt a multi-target tracking algorithm to acquire a video. Therefore, videos acquired by the video monitoring equipment by adopting the multi-target tracking algorithm can be acquired.
When the multi-target tracking algorithm is adopted to collect videos, the video monitoring equipment can respectively establish labels for people and/or objects appearing in the monitored area and track the motion tracks of the people and/or the objects in the monitored area, so that videos marked with the positions of the target frames and the target labels can be formed. In some examples, the object tag may indicate that the object in the object box is a person or object.
After the video shot by the video monitoring device is obtained, video frames can be extracted from the video.
In some examples, the video frames may be extracted randomly from the video, or the video frames may be extracted sequentially in the order of the video frames in the video.
After the video frame is extracted, a pedestrian image including a pedestrian may be acquired from the extracted video frame based on the target label and the position of the target frame of the extracted video frame.
In some examples, since the extracted video frame has the target tags and the positions of the target frames, an image having the same size as the target frames may be acquired from the extracted video frame, and since each target frame has a corresponding target tag, an image in which the target tag is a person may be determined based on the target tags, so that an image of a pedestrian including a pedestrian may be acquired.
In still other examples, after sequentially extracting the video frames in the video, the video sequence may be obtained according to the order of the video frames in the video.
In still other examples, the trajectory analysis may be performed on the video to obtain the trajectory of each pedestrian in the video. And then extracting video sequences corresponding to different pedestrians from the video sequences based on the track of each pedestrian in the video.
For example, the position of each pedestrian in each video frame in the video sequence is determined based on the trajectory of each pedestrian in the video. Then, based on the position of each pedestrian in each video frame, the image containing the pedestrian is extracted from each video frame, and then based on the image of the pedestrian corresponding to each video frame, the video sequence corresponding to the pedestrian can be obtained.
And S230, extracting a target product image from each pedestrian image.
S240, identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image.
And S250, determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
It is understood that steps S230 to S250 in the embodiment of the present invention are similar to or the same as steps S120 to S140 in the first implementation manner of the embodiment of the present invention, and are not described again here.
In some examples, the image set may include at least two pedestrian images of different pedestrians, the pedestrian image of each pedestrian in the image set may be the same, and the pedestrian image of each pedestrian may be 1 or more, for which, on the basis of the method for analyzing the product popularity trend shown in fig. 2, the embodiment of the present invention further provides a possible implementation manner, as shown in fig. 3, which is a flowchart of a third implementation manner of the method for analyzing the product popularity trend according to the embodiment of the present invention, the method may include:
and S310, acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm.
And each video frame in the video is marked with the position of the target frame and the target label.
S320, extracting a plurality of video frames from the video.
And S330, extracting a pedestrian image corresponding to different contained pedestrians from each video frame based on the target label and the position of the target frame of the extracted video frame.
And S340, acquiring a plurality of pedestrian images with the same pedestrian corresponding to the plurality of video frames, and performing quality analysis on the plurality of pedestrian images to obtain quality scores corresponding to the plurality of pedestrian images respectively.
And S350, selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
In some examples, to further improve the accuracy of determining prevalence trends. Multiple video frames may be extracted from the video for each target product. At least one pedestrian may be included in each video frame.
After the plurality of video frames are extracted, a pedestrian image corresponding to a different contained pedestrian may be extracted from each video frame based on the position of the target tag and the target frame of the extracted video frame. Thus, a plurality of pedestrian images corresponding to each video frame can be obtained.
For example, assume that video frame a, video frame B and video frame C are extracted,
assuming that the video frame a contains a pedestrian a and a pedestrian B, the video frame B contains a pedestrian a, a pedestrian B, and a pedestrian C, and the video frame C contains a pedestrian B and a pedestrian C, images of pedestrians corresponding to the contained different pedestrians can be extracted from the video frame a, the video frame B, and the video frame C, respectively. Two pedestrian images including a pedestrian a, three pedestrian images including a pedestrian b, and two pedestrian images including a pedestrian c are obtained.
Then, a plurality of pedestrian images with the same pedestrian can be obtained from a plurality of pedestrian images corresponding to the plurality of video frames, and quality analysis is performed on the plurality of pedestrian images to obtain quality scores corresponding to the plurality of pedestrian images respectively.
In some examples, the purpose of quality analysis of the multiple pedestrian images is to determine a pedestrian image with better quality so as to perform a fashion trend analysis using the pedestrian image with better quality. When the pedestrian in the image is blocked, the image of the pedestrian is fuzzy and darker, or the pedestrian in the image is the side surface, the image quality is poor.
For example, two pedestrian images including the pedestrian a may be subjected to quality analysis, three pedestrian images including the pedestrian b may be subjected to quality analysis, and two pedestrian images including the pedestrian c may be subjected to quality analysis.
After the quality scores of the multiple pedestrian images including the same pedestrian are obtained, the pedestrian image with the highest quality score can be selected from the multiple pedestrian images to serve as the pedestrian image corresponding to the pedestrian.
For example, the pedestrian image with the highest quality score may be selected as the pedestrian image corresponding to the pedestrian a from the two pedestrian images including the pedestrian a, the pedestrian image with the highest quality score may be selected as the pedestrian image corresponding to the pedestrian b from the three pedestrian images including the pedestrian b, or the pedestrian image with the highest quality score may be selected as the pedestrian image corresponding to the pedestrian c from the two pedestrian images including the pedestrian c.
After obtaining the pedestrian image corresponding to each pedestrian, step S360 may be performed to extract the target product image from each pedestrian image.
In some examples, each pedestrian image may include one target product or a plurality of target products, and when one pedestrian image includes a plurality of target products, in a possible implementation manner of the embodiment of the present invention, the pedestrian image may be segmented by using a target product segmentation algorithm to obtain target product images corresponding to the target products at different wearing portions in the pedestrian image. Step S370 may then be performed.
By segmenting the pedestrian image, analysis can be performed only on the basis of the image containing the target product when popular trend analysis is performed, so that the size of the image can be reduced, the recognition rate can be increased, and the efficiency of popular trend analysis can be improved.
And S370, identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image.
And S380, determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
It is understood that steps S310 and S360 to S380 in the embodiment of the present invention are similar to or the same as steps S210 and S230 to S250 in the second implementation manner in the embodiment of the present invention, and are not described again here.
On the basis of the analysis method for the product popularity trend shown in fig. 3, an embodiment of the present invention further provides a possible implementation manner, as shown in fig. 4, which is a flowchart of a fourth implementation manner of the analysis method for the product popularity trend shown in the embodiment of the present invention, and the method may include:
and S410, acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm.
And each video frame in the video is marked with the position of the target frame and the target label.
S420, a plurality of video frames are extracted from the video.
And S430, extracting a pedestrian image corresponding to the contained different pedestrians from each video frame based on the target label and the position of the target frame of the extracted video frame.
And S440, acquiring a plurality of pedestrian images with the same pedestrian corresponding to the plurality of video frames, and performing quality analysis on the plurality of pedestrian images to obtain quality scores corresponding to the plurality of pedestrian images respectively.
And S450, selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
And S460, extracting a target product image from each pedestrian image.
For example, assuming that the target product is a garment, a garment image may be extracted from each pedestrian image. In some examples, the garment may generally include a jacket, trousers, and shoes, and each pedestrian image may be segmented using a garment segmentation algorithm to obtain a jacket image corresponding to the jacket, a trousers image corresponding to the trousers, and a shoes image corresponding to the shoes in the pedestrian image.
S470, performing attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of the target product in each target product image.
For example, the clothing classification algorithm may be used to perform attribute recognition on each image of a jacket to obtain the style and/or color of the jacket in each image of the jacket, the clothing classification algorithm may be used to perform attribute recognition on each image of trousers to obtain the style and/or color of the trousers in each image of the trousers, and the clothing classification algorithm may be used to perform attribute recognition on each image of shoes to obtain the style and/or color of the shoes in each image of the shoes.
S480, analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
In some examples, when identifying the attribute value of the visual attribute of the target product in the target product image, the attribute identification may be performed on each target product image by using a target product classification algorithm, so that the visual attribute of the target product in each target product image may be obtained.
In still other examples, the target product classification algorithm may be a pre-trained neural network algorithm. The neural network algorithm obtained by pre-training can be obtained by training by adopting a product image sample labeled with visual attributes.
In still other examples, a corresponding target product classification algorithm may be set for each target product, so that for the target product, the target product classification algorithm corresponding to the target product is used to determine the visual attribute of the target product.
By adopting the target product classification algorithm corresponding to each target product, the determined visual attribute can be more accurate, so that the accuracy of the subsequently determined popular trend can be improved.
After the visual attributes of the target products in each target product image are obtained, the visual attributes of the target products in all the target product images in the image set can be analyzed, and the attribute values of the visual attributes of the target products in each target product image are obtained.
In some examples, for each visual attribute, a proportion of the visual attribute in the visual attribute of the target product in all of the images of the target product in the image set may be calculated and then used as the attribute value of the visual attribute.
For example, for each style, the ratio of the style to the styles of the target products in all the images of the target products in the image set may be calculated, and then the ratio may be used as the attribute value of the style. Then, through step S490, the most popular style or more popular styles are determined.
For another example, for each color, the ratio of the color to the color of the target product in all the target product images in the image set may be calculated, and then the ratio may be used as the attribute value of the color. The most popular colors or more popular colors are then determined, via step S490.
For example, the style and/or color of the upper garment in all the upper garment images in the image set may be analyzed to obtain the attribute value of the style and/or color of the upper garment in each upper garment image, the style and/or color of the trousers in all the trousers images in the image set may be analyzed to obtain the attribute value of the style and/or color of the trousers in each trousers image, and the style and/or color of the shoes in all the shoe images in the image set may be analyzed to obtain the attribute value of the style and/or color of the shoes in each shoe image.
And S490, determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
For example, the fashion trend and/or the color trend of the jacket may be determined based on the attribute value of the style and/or the color of the jacket in each jacket image, the fashion trend and/or the color trend of the trousers may be determined based on the attribute value of the style and/or the color of the trousers in each trousers image, and the fashion trend and/or the color trend of the shoes may be determined based on the attribute value of the style and/or the color of the shoes in each shoe image.
It is understood that steps S410 to S460, S490 in the embodiment of the present invention are similar to or the same as steps S310 to S360, S380 in the third implementation manner of the embodiment of the present invention, and are not described again here.
Corresponding to the above method embodiment, an embodiment of the present invention further provides an apparatus for analyzing a product popularity trend, as shown in fig. 5, which is a schematic structural diagram of an apparatus for analyzing a product popularity trend according to an embodiment of the present invention, and the apparatus may include:
the image set obtaining module 510 is configured to obtain an image set of pedestrian images collected by the video monitoring device, where pedestrians in at least two pedestrian images in the image set are different, a ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set include target product images;
an image extraction module 520 for extracting a target product image from each pedestrian image;
an attribute value identification module 530, configured to identify each target product image, and obtain an attribute value of each visual attribute of a target product in each target product image;
and the popularity trend analysis module 540 is used for determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
When determining the popularity trend of the target product, the device for analyzing the popularity trend of the product provided by the embodiment of the invention can firstly acquire the image set of the pedestrian images acquired by the monitoring equipment, then extract the image of the target product from each image of the pedestrian, identify each image of the target product, obtain the attribute value of each visual attribute of the target product in each image of the target product, and finally determine the popularity trend of the target product based on the attribute value of each visual attribute of the target product in each image of the target product. Because the pedestrians in at least two pedestrian images in the image set are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images in the image set are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and can represent the wearing tendency of the pedestrians, the popular tendency of the target products is analyzed based on the target products in the pedestrian images, the popular tendency analysis by using a data set containing the goods change data and/or the goods return data can be avoided, and the accuracy of the popular tendency analysis of the products is improved.
In some examples, the image set acquisition module 510 may include:
the video acquisition submodule is used for acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm, wherein each video frame in the video is marked with a position of a target frame and a target label;
and the image set acquisition module submodule is used for extracting the video frames from the video and acquiring pedestrian images containing pedestrians from the extracted video frames on the basis of the target labels and the positions of the target frames of the extracted video frames.
In still other examples, the image set acquisition module sub-module includes:
a video frame extraction unit for extracting a plurality of video frames from a video;
an image segmentation unit for extracting, for each video frame, pedestrian images corresponding to different contained pedestrians from the video frame based on the extracted target tags and positions of the target frames of the video frame;
the quality analysis unit is used for acquiring a plurality of pedestrian images which are corresponding to the video frames and have the same pedestrian, and performing quality analysis on the pedestrian images to obtain quality scores which respectively correspond to the pedestrian images;
and the image set acquisition unit is used for selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
In some examples, the image extraction module 520 is specifically configured to:
segmenting the pedestrian image by adopting a target product segmentation algorithm to obtain target product images corresponding to target products at different wearing parts in the pedestrian image;
in some examples, attribute value identification module 530 includes:
the visual attribute identification submodule is used for carrying out attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of the target product in each target product image;
and the attribute value analysis submodule is used for analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
In some examples, popularity trend analysis module 540 is specifically configured to:
determining a target product corresponding to the visual attribute with the maximum attribute value as the most popular product based on the attribute values of the visual attributes; and/or
And determining that the target product corresponding to the visual attribute with the attribute value sequence positioned before the preset digit is a popular product based on the sequence of the attribute values of the visual attributes from large to small.
An embodiment of the present invention further provides an electronic device, as shown in fig. 6, including a processor 601, a communication interface 602, a memory 603, and a communication bus 604, where the processor 601, the communication interface 602, and the memory 603 complete mutual communication through the communication bus 604,
a memory 603 for storing a computer program;
the processor 601, when executing the program stored in the memory 603, implements the steps of any of the above methods for analyzing popularity of products, for example, the following steps may be implemented:
acquiring an image set of pedestrian images acquired by video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
extracting a target product image from each pedestrian image;
identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image;
and determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
When determining the popularity trend of the target product, the electronic device provided by the embodiment of the invention can firstly acquire the image set of the pedestrian images acquired by the monitoring device, then extract the target product image from each pedestrian image, identify each target product image, obtain the attribute value of each visual attribute of the target product in each target product image, and finally determine the popularity trend of the target product based on the attribute value of each visual attribute of the target product in each target product image. Because the pedestrians in at least two pedestrian images in the image set are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images in the image set are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and can represent the wearing tendency of the pedestrians, the popular tendency of the target products is analyzed based on the target products in the pedestrian images, the popular tendency analysis by using a data set containing the goods change data and/or the goods return data can be avoided, and the accuracy of the popular tendency analysis of the products is improved.
The communication bus mentioned in the electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.
The communication interface is used for communication between the electronic equipment and other equipment.
The Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the method for analyzing the popularity trend of any product described above is implemented, for example, the following steps may be implemented:
acquiring an image set of pedestrian images acquired by video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
extracting a target product image from each pedestrian image;
identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image;
and determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
When determining the popularity trend of the target product, the computer-readable storage medium provided by the embodiment of the invention may first obtain an image set of pedestrian images acquired by a monitoring device, then extract the target product image from each pedestrian image, identify each target product image, obtain attribute values of each visual attribute of the target product in each target product image, and finally determine the popularity trend of the target product based on the attribute values of each visual attribute of the target product in each target product image. Because the pedestrians in at least two pedestrian images in the image set are different and the ratio of the number of the pedestrian images with different pedestrians is within the preset ratio range, the pedestrians in the different pedestrian images in the image set are not completely repeated, and because the products in the pedestrian images are worn on the pedestrians and can represent the wearing tendency of the pedestrians, the popular tendency of the target products is analyzed based on the target products in the pedestrian images, the popular tendency analysis by using a data set containing the goods change data and/or the goods return data can be avoided, and the accuracy of the popular tendency analysis of the products is improved.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (14)
1. A method for analyzing a product popularity trend, the method comprising:
acquiring an image set of pedestrian images acquired by video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
extracting a target product image from each pedestrian image;
identifying each target product image to obtain attribute values of each visual attribute of a target product in each target product image;
and determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
2. The method of claim 1, wherein said obtaining an image set of pedestrian images captured by a video surveillance device comprises:
acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm, wherein each video frame in the video is marked with a position and a target label of a target frame;
and extracting video frames from the video, and acquiring pedestrian images containing pedestrians from the extracted video frames based on the target labels and the positions of the target frames of the extracted video frames.
3. The method according to claim 2, wherein the extracting a video frame from the video and acquiring a pedestrian image including a pedestrian from the extracted video frame based on the target label and the position of the target frame of the extracted video frame comprises:
extracting a plurality of video frames from the video;
for each video frame, extracting a pedestrian image corresponding to different contained pedestrians from the video frame based on the target label and the position of the target frame of the extracted video frame;
acquiring a plurality of pedestrian images with the same pedestrian corresponding to the plurality of video frames, and performing quality analysis on the plurality of pedestrian images to obtain quality scores corresponding to the plurality of pedestrian images respectively;
and selecting the pedestrian image with the highest quality score from the plurality of pedestrian images as the pedestrian image corresponding to the pedestrian.
4. The method of claim 1, wherein said extracting a target product image from each of said pedestrian images comprises:
and segmenting the pedestrian image by adopting a target product segmentation algorithm to obtain target product images corresponding to target products of different wearing parts in the pedestrian image.
5. The method of claim 1, wherein the identifying each target product image to obtain the attribute value of each visual attribute of the target product in the each target product image comprises:
performing attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of a target product in each target product image;
and analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
6. The method of claim 1, wherein determining a popularity trend of the target product based on the attribute values of the respective visual attributes of the target product in each target product image comprises:
based on the attribute values of the visual attributes, determining a target product corresponding to the visual attribute with the maximum attribute value as a most popular product; and/or
And determining that the target product corresponding to the visual attribute with the attribute value sequence positioned before the preset digit is a popular product based on the sequence of the attribute values of the visual attributes from large to small.
7. An apparatus for analyzing a product popularity trend, the apparatus comprising:
the image set acquisition module is used for acquiring an image set of pedestrian images acquired by video monitoring equipment, wherein pedestrians in at least two pedestrian images in the image set are different, the ratio of the number of the pedestrian images with different pedestrians is within a preset ratio range, and the pedestrian images in the image set comprise target product images;
the image extraction module is used for extracting a target product image from each pedestrian image;
the attribute value identification module is used for identifying each target product image to obtain the attribute value of each visual attribute of the target product in each target product image;
and the popularity trend analysis module is used for determining the popularity trend of the target product based on the attribute values of the visual attributes of the target product in each target product image.
8. The apparatus of claim 7, wherein the image set acquisition module comprises:
the video acquisition submodule is used for acquiring a video acquired by the video monitoring equipment by adopting a multi-target tracking algorithm, wherein each video frame in the video is marked with a position of a target frame and a target label;
and the image set acquisition module submodule is used for extracting a video frame from the video and acquiring a pedestrian image containing a pedestrian from the extracted video frame on the basis of the position of the target label and the target frame of the extracted video frame.
9. The apparatus of claim 8, wherein said image set acquisition module submodule comprises:
a video frame extraction unit for extracting a plurality of video frames from the video;
an image segmentation unit, configured to extract, for each of the video frames, a pedestrian image corresponding to a different contained pedestrian from the video frame based on the target label and the position of the target frame of the extracted video frame;
the quality analysis unit is used for acquiring a plurality of pedestrian images which are corresponding to the video frames and have the same pedestrian, and performing quality analysis on the pedestrian images to obtain quality scores which respectively correspond to the pedestrian images;
and the image set acquisition unit is used for selecting the pedestrian image with the highest quality score from the multiple pedestrian images as the pedestrian image corresponding to the pedestrian.
10. The apparatus of claim 7, wherein the image extraction module is specifically configured to:
and segmenting the pedestrian image by adopting a target product segmentation algorithm to obtain target product images corresponding to target products of different wearing parts in the pedestrian image.
11. The apparatus of claim 7, wherein the attribute value identification module comprises:
the visual attribute identification submodule is used for carrying out attribute identification on each target product image by adopting a target product classification algorithm to obtain the visual attribute of a target product in each target product image;
and the attribute value analysis submodule is used for analyzing the visual attributes of the target products in all the target product images in the image set to obtain the attribute values of the visual attributes of the target products in each target product image.
12. The apparatus of claim 7, wherein the popularity trend analysis module is specifically configured to:
based on the attribute values of the visual attributes, determining a target product corresponding to the visual attribute with the maximum attribute value as a most popular product; and/or
And determining that the target product corresponding to the visual attribute with the attribute value sequence positioned before the preset digit is a popular product based on the sequence of the attribute values of the visual attributes from large to small.
13. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any of claims 1-6 when executing a program stored in the memory.
14. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911102019.6A CN112801681A (en) | 2019-11-12 | 2019-11-12 | Product popularity trend analysis method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911102019.6A CN112801681A (en) | 2019-11-12 | 2019-11-12 | Product popularity trend analysis method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112801681A true CN112801681A (en) | 2021-05-14 |
Family
ID=75802975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911102019.6A Pending CN112801681A (en) | 2019-11-12 | 2019-11-12 | Product popularity trend analysis method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112801681A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268544A (en) * | 2014-10-14 | 2015-01-07 | 陶晨 | Evaluation system for clothes visual effect |
| CN107766861A (en) * | 2017-11-14 | 2018-03-06 | 深圳码隆科技有限公司 | The recognition methods of character image clothing color, device and electronic equipment |
| CN108960499A (en) * | 2018-06-27 | 2018-12-07 | 东华大学 | A kind of Fashion trend predicting system merging vision and non-vision feature |
| CN111145242A (en) * | 2019-12-26 | 2020-05-12 | 赛维时代科技股份有限公司 | Method, smart device, and computer-readable storage medium for predicting popularity trend |
-
2019
- 2019-11-12 CN CN201911102019.6A patent/CN112801681A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268544A (en) * | 2014-10-14 | 2015-01-07 | 陶晨 | Evaluation system for clothes visual effect |
| CN107766861A (en) * | 2017-11-14 | 2018-03-06 | 深圳码隆科技有限公司 | The recognition methods of character image clothing color, device and electronic equipment |
| CN108960499A (en) * | 2018-06-27 | 2018-12-07 | 东华大学 | A kind of Fashion trend predicting system merging vision and non-vision feature |
| CN111145242A (en) * | 2019-12-26 | 2020-05-12 | 赛维时代科技股份有限公司 | Method, smart device, and computer-readable storage medium for predicting popularity trend |
Non-Patent Citations (2)
| Title |
|---|
| 傅白璐: "基于数据挖掘的女装廓形流行趋势智能分析预测系统研究", 《中国优秀博士学位论文全文数据库工程科技I辑》, pages 11 - 20 * |
| 胡江华: "静态图像的行人分割及其应用", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, pages 24 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9996737B2 (en) | Method and system for automatically recognizing facial expressions via algorithmic periocular localization | |
| CN108549846B (en) | Pedestrian detection and statistics method combining motion characteristics and head-shoulder structure | |
| Kalsotra et al. | Background subtraction for moving object detection: explorations of recent developments and challenges | |
| CN109727275B (en) | Object detection method, device, system and computer readable storage medium | |
| CN109670383B (en) | Video shielding area selection method and device, electronic equipment and system | |
| Ma et al. | Counting people crossing a line using integer programming and local features | |
| CN112417970A (en) | Target object identification method, device and electronic system | |
| CN110610123A (en) | Multi-target vehicle detection method and device, electronic equipment and storage medium | |
| Yamamoto et al. | Multiple players tracking and identification using group detection and player number recognition in sports video | |
| CN115830002A (en) | Infrared image quality evaluation method and device | |
| CN110929668A (en) | Commodity detection method and device based on unmanned goods shelf | |
| CN109815902B (en) | Method, device and equipment for acquiring pedestrian attribute region information | |
| Liu et al. | Deep learning-based automatic player identification and logging in american football videos | |
| CN112801681A (en) | Product popularity trend analysis method and device, electronic equipment and storage medium | |
| Frontoni et al. | People counting in crowded environment and re-identification | |
| CN115482569A (en) | Target passenger flow statistical method, electronic device and computer readable storage medium | |
| CN112069357B (en) | Video resource processing method and device, electronic equipment and storage medium | |
| Tagore et al. | SMSNet: A Novel Multi-scale Siamese Model for Person Re-Identification. | |
| Nyström | Evaluation of Multiple Object Tracking in Surveillance Video | |
| Barbuzza et al. | A shadow removal approach for a background subtraction algorithm | |
| Nguyen et al. | Video monitoring system: Counting people by tracking | |
| CN111476132A (en) | Video scene recognition method and device, electronic equipment and storage medium | |
| Mohaghegh et al. | A four-component people identification and counting system using deep neural network | |
| Parisot et al. | Training with corrupted labels to reinforce a probably correct teamsport player detector | |
| CN112308023B (en) | Method, device and equipment for identifying pedestrian movement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |