CN111695488B - Method, device, equipment and storage medium for identifying interest surface - Google Patents

Method, device, equipment and storage medium for identifying interest surface Download PDF

Info

Publication number
CN111695488B
CN111695488B CN202010516452.0A CN202010516452A CN111695488B CN 111695488 B CN111695488 B CN 111695488B CN 202010516452 A CN202010516452 A CN 202010516452A CN 111695488 B CN111695488 B CN 111695488B
Authority
CN
China
Prior art keywords
target
interest
target area
information
grid
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.)
Active
Application number
CN202010516452.0A
Other languages
Chinese (zh)
Other versions
CN111695488A (en
Inventor
路新江
熊辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Baidu Netcom Science and Technology Co Ltd
Original Assignee
Beijing Baidu Netcom Science and Technology Co Ltd
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 Beijing Baidu Netcom Science and Technology Co Ltd filed Critical Beijing Baidu Netcom Science and Technology Co Ltd
Priority to CN202010516452.0A priority Critical patent/CN111695488B/en
Publication of CN111695488A publication Critical patent/CN111695488A/en
Application granted granted Critical
Publication of CN111695488B publication Critical patent/CN111695488B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/176—Urban or other man-made structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • 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
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0639—Item locations
    • 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
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Evolutionary Computation (AREA)
  • Finance (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Databases & Information Systems (AREA)
  • Mathematical Physics (AREA)
  • Multimedia (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Evolutionary Biology (AREA)
  • Image Processing (AREA)

Abstract

The application discloses a method, a device, equipment and a storage medium for identifying an interest surface, and relates to artificial intelligence and deep learning. The specific implementation scheme is as follows: acquiring target function category information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area; acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model. The application realizes the identification of the interest surface in the target area by utilizing the computer vision technology, does not need to carry out manual labeling based on street view images, can reduce the cost, improve the efficiency, has higher accuracy, can be suitable for different scenes, has larger application range, can be applied to the identification of massive interest surfaces, and further enables the description of the surface-shaped information of massive interest points to be possible.

Description

Method, device, equipment and storage medium for identifying interest surface
Technical Field
The embodiment of the application relates to artificial intelligence and deep learning in computer technology, in particular to a method, a device, equipment and a storage medium for identifying an interest surface.
Background
An Area of Interest (AOI) refers to an Area of an electronic map that is actually present in the physical world, and is used to identify urban functions (such as schools, communities, hospitals, shopping centers, etc.) and geographic boundaries thereof represented by the geographic location. The interest surface can vividly depict the geographic boundary attribute of the city function; in addition, in the task of completing the labels of the urban interest points (Point of Interest, POIs), when the geographical boundary of the interest surface is known, the data such as pictures, characters, tracks and the like generated by the user in the geographical range covered by the interest surface can be associated with the relevant interest points, so that semantic label information of the interest points is enriched; the interest level is also helpful to judge whether the user visits the interest point, so that accurate shop recommendation can be performed.
In the prior art, professional acquisition equipment is required to acquire street view images, such as acquisition vehicles and the like, and professional acquisition personnel perform manual labeling, so that the outline of the interest surface is labeled; or clustering the interest points of the same functional category through a clustering algorithm by a bottom-up clustering method so as to determine the outline of the interest surface.
The manual collection mode in the prior art has the advantages of higher equipment cost and labor cost, low efficiency and incapability of being applied to the identification of massive interest surfaces; the bottom-up clustering method is complex in process and cannot guarantee accuracy.
Disclosure of Invention
The application provides a method, a device, equipment and a storage medium for identifying an interest surface, which are used for reducing the cost of identifying the interest surface and improving the identification efficiency and accuracy.
According to a first aspect of the present application, there is provided a method for identifying an interest surface, including:
acquiring target function class information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area;
acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area;
and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model.
According to a second aspect of the present application, there is provided an apparatus for face recognition, comprising:
the acquisition module is used for acquiring target function category information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area;
The processing module is used for acquiring corresponding input data of a preset computer vision model according to the target function category information and the geographic attribute related data of the target area;
and the identification module is used for identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model.
According to a third aspect of the present application, there is provided an electronic device comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein, the liquid crystal display device comprises a liquid crystal display device,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of the first aspect.
According to a fourth aspect of the present application there is provided a non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of the first aspect.
According to a fifth aspect of the present application there is provided a computer program comprising program code which, when run by a computer, performs the method according to the first aspect.
According to a sixth aspect of the present application, there is provided a computer program product comprising: a computer program stored in a readable storage medium, from which it can be read by at least one processor of an electronic device, the at least one processor executing the computer program causing the electronic device to perform the method of the first aspect.
According to the method, the device, the equipment and the storage medium for identifying the interest surface, the target function category information corresponding to the target interest surface in the target area to be identified and the geographic attribute related data of the target area are obtained; acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model. The application realizes the identification of the interest surface in the target area by utilizing the computer vision technology, does not need to carry out manual labeling based on street view images, can reduce the cost, improve the efficiency, has higher accuracy, can be suitable for different scenes, has larger application range, can be applied to the identification of massive interest surfaces, and further enables the description of the surface-shaped information of massive interest points to be possible.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.
Drawings
The drawings are included to provide a better understanding of the present application and are not to be construed as limiting the application. Wherein:
FIG. 1 is a schematic view of a scene of a method for identifying an interest surface according to an embodiment of the present application;
FIG. 2 is a flow chart of a method for identifying an interest surface according to an embodiment of the present application;
FIG. 3 is a flow chart of a method for identifying an interest surface according to an embodiment of the present application;
FIG. 4A is a satellite image of a target area provided in accordance with one embodiment of the present application;
FIG. 4B is a satellite image marked with a point of interest identification provided in accordance with an embodiment of the present application;
FIG. 4C is a satellite image marked with a point of interest identification provided in accordance with another embodiment of the present application;
FIG. 4D is a schematic diagram of results obtained by an image instance segmentation model according to an embodiment of the present application;
FIG. 5 is a flow chart of a method for identifying an interest surface according to an embodiment of the present application;
FIG. 6 is a schematic diagram of results obtained by an image semantic segmentation model according to an embodiment of the present application;
FIG. 7 is a schematic diagram of an image semantic segmentation model provided according to an embodiment of the present application;
FIG. 8 is a flow chart of a method for identifying an interest surface according to an embodiment of the present application;
FIG. 9 is a block diagram of an interest surface recognition device provided in accordance with an embodiment of the present application;
FIG. 10 is a block diagram of an electronic device for implementing a method of interest surface identification in accordance with an embodiment of the present application.
Detailed Description
Exemplary embodiments of the present application will now be described with reference to the accompanying drawings, in which various details of the embodiments of the present application are included to facilitate understanding, and are to be considered merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
In the prior art, manual labeling is performed based on street view images to obtain the outline of the interest surface, professional equipment and professional staff are needed, equipment cost and labor cost are high, efficiency is low, and the bottom-up clustering method is complex in process and cannot guarantee accuracy. That is, the method for identifying the interest surfaces in the prior art has high cost and low identification efficiency and accuracy, and cannot be applied to the identification of massive interest surfaces.
Aiming at the problems existing in the prior art, the application considers that the identification of the interest surface in the target area is realized by utilizing the computer vision technology, and the application can reduce the cost, improve the efficiency and have higher accuracy by acquiring the target function category information corresponding to the target interest surface in the target area to be identified and the geographic attribute related data of the target area without manually marking based on street view images, thereby being applicable to different scenes, having larger application range and being applicable to the identification of massive interest surfaces and further enabling the description of the planar information of massive interest points.
The method for identifying the interest surface provided by the application can be applied to a scene shown in fig. 1, wherein the database 11 is used for storing the geographic attribute related data of different areas, the electronic equipment 10 can acquire the geographic attribute related data of the target area to be identified from the database 11, in addition, the electronic equipment 10 can also acquire the target function type information corresponding to the target interest surface in the target area, for example, the target function type information can be preset by a user, further, input data of a corresponding preset computer vision model is acquired according to the target function type information and the geographic attribute related data of the target area, and then the outline of the target interest surface in the target area is identified according to the input data and the corresponding preset computer vision model.
Alternatively, after the contour of the target interest surface is identified, the contour of the target interest surface may be output through the display 13. Further, geographic boundary attributes of the city function can be described through the interest surface; in addition, in the task of completing the labels of the urban interest points, when the geographical boundary of the interest surface is known, the data such as pictures, characters, tracks and the like generated by the user in the geographical range covered by the interest surface can be associated with the relevant interest points, so that semantic label information of the interest points is enriched; the interest level is also helpful to judge whether the user visits the interest point, so that accurate shop recommendation can be performed.
The process of identifying the interest surface provided by the application will be described in detail with reference to specific embodiments.
An embodiment of the application provides a method for identifying an interest surface, and fig. 2 is a flowchart of the method for identifying an interest surface provided by the embodiment of the application. The execution body may be the electronic device 10 in fig. 1, as shown in fig. 2, and the method specifically includes the following steps:
s201, obtaining target function category information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area.
In this embodiment, the target city may be first divided into a plurality of regions according to road network information, and optionally, the regions may be divided according to road networks classified by preset roads, for example, the city may be divided into a plurality of blocks by seven-level roads through which the motor vehicle may pass, so that semantic integrity of each region may be ensured.
When the interest surface identification is required to be carried out on one or more areas in the plurality of areas of the target city, the target area can be selected by a user, and after the target area selection instruction of the user is received, the target area to be identified is determined from the plurality of areas according to the target area selection instruction.
Further, after the target area to be identified is determined, target function category information corresponding to the target interest surface in the target area and geographic attribute related data of the target area can be obtained.
The target function class information may be preset by a user, for example, a target area may include geographic entities of multiple function classes, such as schools, residential areas, hospitals, and the like, and when the user needs to identify the outline of the interest plane of the function class of the school in the target area, the user may set the target function class information corresponding to the target interest plane as the school.
The geographic attribute related data of the target area is data for characterizing the geographic attribute of the target area, which may include, but is not limited to, satellite images of the target area, point of interest (Point of Information, POI) information in the target area, road network information of the target area, user behavior information in the target area, etc., and the geographic attribute related data of the target area may be determined according to the difficulty level of data acquisition, such as some underdeveloped areas or emerging areas without satellite images, so that other geographic attribute related data may be used. Of course, according to the difference of the geographic attribute related data of the target area, a corresponding preset computer vision model can be selected to perform subsequent contour recognition processing of the target interest surface.
For example, the geographic attribute related data of the target area may include satellite images of the target area and interest point information of the target area, and the preset computer vision model is an image instance segmentation model; for another example, the geographic attribute related data of the target area may include feature information of points of interest of the target area and feature information of road networks of the target area, and may also include user behavior information in the target area, etc., where the preset computer vision model is an image semantic segmentation model.
Of course, the preset computer vision model in this embodiment is not limited to the above-listed models, and the preset computer vision model may be other models, and the geographic attribute related data of the target area may be determined according to the model requirements. In addition, when a preset computer vision model for carrying out contour recognition processing on the target interest surface is determined, the geographic attribute related data of the target area can be determined according to the preset computer vision model.
S202, acquiring input data of a corresponding preset computer vision model according to the target function category information and the geographic attribute related data of the target area.
In this embodiment, since the contour recognition processing of the target interest surface is performed by the preset computer vision model, the input data of the preset computer vision model needs to be obtained according to the obtained target function class information and the geographic attribute related data of the target area.
For example, the geographic attribute related data of the target area may include a satellite image of the target area and interest point information of the target area, the preset computer vision model is an image instance segmentation model, and the input data may be a satellite image marked with interest point identification; for another example, the geographic attribute related data of the target area may include feature information of points of interest of the target area and road network feature information of the target area, and may also include user behavior information in the target area, etc., and the preset computer vision model is an image semantic segmentation model, and the input data may be a semantic feature map characterizing the target area obtained according to the geographic attribute related data. The preset computer vision model in this embodiment is not limited to the above listed models, and thus the input data and the process of obtaining the input data can be determined according to the preset computer vision model, which is not described here again.
S203, identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model.
In this embodiment, input data is input into a preset computer vision model, and the preset computer vision model uses a corresponding computer vision technology to obtain the outline of the target interest surface in the target area after processing the input data.
It will be appreciated that after the contours of the target interest surface are identified, the contours of the target interest surface may be output via a display screen.
According to the interest surface identification method provided by the embodiment, the target function category information corresponding to the target interest surface in the target area to be identified and the geographic attribute related data of the target area are obtained; acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model. In the embodiment, the identification of the interest surface in the target area is realized by utilizing a computer vision technology, and the manual labeling is not needed based on the street view image, so that the cost can be reduced, the efficiency is improved, the accuracy is higher, the method and the device are applicable to different scenes, the application range is wider, and the method and the device can be applied to the identification of massive interest surfaces, so that the description of the planar information of massive interest points is possible.
FIG. 3 is a flowchart of a method for identifying an interest surface according to another embodiment of the present invention. On the basis of the above embodiment, the present embodiment provides an interest surface recognition method, which is directed to the case where satellite images of a target area are available.
In this embodiment, since the satellite image contains abundant physical space information, such as buildings, streets, vegetation, etc., which are good for computer vision technology, the problem of interest surface identification can be analogized with the problem of image instance segmentation, which is to identify objects in the image and delineate object boundaries, whereas in this embodiment, the image instance segmentation technique is used to identify interest surfaces in the satellite image and delineate boundaries of interest surfaces. Considering that the interest surface is closely related to the city function, but the semantic information about the city function cannot be reflected from the satellite image, the interest surface cannot be effectively identified by only relying on the satellite image, so that the semantic information of the point of interest (POI) is introduced into the satellite image, and the satellite image is enhanced as an input of an image instance segmentation model in the embodiment.
Thus, the geographic attribute related data of the target area in the present embodiment may include satellite images of the target area and point of interest information of the target area; the preset computer vision model is an image instance segmentation model.
The method for identifying the interest surface provided by the embodiment comprises the following specific steps:
S301, obtaining target function category information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area.
The geographic attribute related data of the target area may include satellite images of the target area and interest point information of the target area.
In this embodiment, the principle and implementation of S301 are similar to those of S201 described above, and will not be repeated here.
Further, S202 in the above embodiment may specifically include the following S302-S303, which specifically includes the following procedures:
s302, screening a set of first target interest points which are the same as the target function category in the target area according to the target function category information and the interest point information of the target area.
In this embodiment, for the interest points in the target area, screening may be performed according to the target function class information corresponding to the target interest surface, and the first target interest point identical to the target function class may be screened from the interest points in the target area, so as to obtain the set of the first target interest points.
S303, marking the identification of the first target interest point on the corresponding position of the satellite image of the target area according to the set of the first target interest point, and obtaining the satellite image marked with the interest point identification as input data of the image instance segmentation model.
In this embodiment, since the satellite image does not have the semantic information of the city function, and the interest points have the semantic information of the city function, marking the first target interest points on the satellite image can enable the satellite image to have the semantic information of the city function, that is, how many first target interest points exist in the target area and the positions, the distributions and the like of the first target interest points can be known through the satellite image marked with the interest point marks. The satellite image of the target area shown in fig. 4A is marked to obtain the satellite image marked with the interest point mark shown in fig. 4B, where the dot is the mark of the first target interest point.
Optionally, when the identifier of the first target interest point is marked on the corresponding position of the satellite image of the target area, the first target interest point may be selected from the set of first target interest points according to a random sequence, and if the first target interest point is not blocked by the marked interest point identifier in the satellite image, the first target interest point is marked on the corresponding position of the satellite image of the target area.
In this embodiment, considering that the interest point information has great differences in different areas, how to selectively and reasonably mark the interest point identification on the satellite image is important to accurately and effectively identify the target interest plane. Therefore, in this embodiment, the first target point of interest is selected from the set of first target points of interest to be marked by using a predetermined rule, and specifically, the predetermined rule may be: each point of interest corresponds to an identifier (icon); the interest points of different functional categories adopt different identifications; the category of the interest points marked by the satellite image is the same as the target function category corresponding to the target interest surface, and each interest point mark (icon) is not shielded; and selecting the interest points in a random order for marking until no interest points can be marked. In this embodiment, a heuristic algorithm may be used to mark the point of interest identifier on the satellite image according to the above rule. Specifically, a first target interest point is randomly selected from a first target interest point set, an identification is marked on a satellite image according to the position information of the first target interest point, then the next first target interest point is randomly selected, whether the first target interest point is blocked with the marked interest point identification is judged, if not, the first target interest point is marked, if not, the first target interest point is discarded, and the like until no first target interest point which can be continuously marked on the satellite image exists in the first target interest point set. In this embodiment, the random order is adopted to avoid deviation of the identification result of the interest surface caused by possible uneven distribution of the interest points. For the satellite image shown in fig. 4A, through the selective marking process described above, it can be obtained that the identification distribution of the first target interest point is relatively uniform and no occlusion exists as shown in fig. 4C.
Further, S203 in the foregoing embodiment may specifically include S304 as follows:
s304, image instance segmentation is carried out on the satellite image marked with the interest point mark by the image instance segmentation model, and the outline of the target interest surface in the target area is identified.
In this embodiment, the image instance segmentation model is an existing neural network model, such as Mask R-CNN, and the image instance segmentation process is performed on the satellite image marked with the interest point identifier by using the image instance segmentation model, which is similar to the image instance segmentation process performed by using the existing image instance segmentation model, only the input data is different, and the specific process is not described here again. As an example, the image instance segmentation model output result is shown in fig. 4D, where the outline of the target interest surface within the target region (white bold line range) is a white thin line range.
It should be noted that, the image instance segmentation model in this embodiment may be trained in advance through training data, where the training data may be a satellite image marked with a point of interest identifier, and the contour of the target interest surface has been identified, and the training process is not described herein.
FIG. 5 is a flowchart of a method for identifying an interest surface according to another embodiment of the present invention. Based on the above embodiments, the present embodiment provides a method for identifying an interest plane, which is aimed at a situation where a satellite image of a target area is not available, for example, a less developed area, or an emerging area, or other areas where it is difficult to acquire a satellite image.
In this embodiment, since it is difficult to acquire a satellite image of a target area, the point of interest characteristic information of the target area and the road network characteristic information of the target area, which are cheaper and easier to acquire, are used instead as the geographic attribute related data of the target area. In this embodiment, the preset computer vision model is an image semantic segmentation model.
As shown in fig. 5, the method for identifying an interest surface provided in this embodiment specifically includes the following steps:
s401, obtaining target function category information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area.
The geographic attribute related data of the target area may include feature information of points of interest of the target area and feature information of road networks of the target area.
Optionally, the feature information of the interest point of the target area may include feature information of three dimensions: the functional category distribution of the interest points in the target area, the functional category distribution of the adjacent interest points of each interest point, the position of each interest point, and other characteristic information such as the names of the interest points can be included. And the road network characteristic information of the target area may include, but is not limited to, a position of the road network, a start point and an end point of the road, a length of the road, a density of the road network, a distance between the interest point and the road, and the like.
In this embodiment, the principle and implementation of S401 are similar to those of S201 described above, and will not be repeated here.
Further, S202 in the above embodiment may specifically include S402 to S404 as follows:
s402, determining the center position of the target interest surface according to the interest point characteristic information of the target area and the road network characteristic information of the target area.
In this embodiment, a set of second target points of interest that are the same as the target function class may be selected from the points of interest of the target area according to the target function class information; and determining a representative interest point from the second target interest point set according to the interest point characteristic information of each second target interest point in the second target interest point set, the road network characteristic information of the target area and a preset classification model, and taking the position of the representative interest point as the center position of the target interest surface. Specifically, the interest point feature information of each second target interest point and the road network feature information of the target area are sequentially input into a preset classification model, and whether the interest point of each second target interest point is the center position of the target interest surface is sequentially judged until the interest point of a certain second target interest point is determined to be the center position of the target interest surface. The preset classification model in this embodiment may be any classification model.
S403, gridding the target area, and constructing a plurality of feature graphs of each grid according to the feature information of the interest points falling into the grid and the feature information of the road network in the grid.
In this embodiment, the target area is first gridded, that is, the target area is divided into a plurality of grids, then, for any one grid, the interest points falling into the grid are determined, and then, the interest point feature information and the road network feature information falling into the grid are obtained, and a plurality of feature maps of the grid are constructed, for example, the feature maps of the road network and the interest point feature map may be included, and the feature map of the interest point may specifically include a feature map of an interest point function category, a feature map of an interest point name, a feature map of adjacent interest point information, and so on.
S404, taking the central position of the target interest surface and the characteristic diagrams of each grid as input data of the image semantic segmentation model.
In this embodiment, the center position of the target interest surface obtained in the above step and the multiple feature maps of each grid are input into the image semantic segmentation model to perform subsequent recognition of the target interest surface, and the image semantic segmentation processing can be performed on each grid through gridding, so that the accuracy and efficiency of the image semantic segmentation processing can be improved, and the determined contour of the target interest surface can be more accurate.
Further, S203 in the above embodiment may specifically include the following S405 to S406, which specifically includes the following procedures:
s405, sequentially judging whether grids around the central position of the target interest surface belong to the target interest surface or not by using the central position of the target interest surface as the center through the image semantic segmentation model;
s406, determining the outline of the target interest surface in the target area according to the grid positions belonging to the target interest surface.
In this embodiment, whether surrounding grids belong to the target interest surface is sequentially determined by taking the center position of the target interest surface as the center, if so, the next grid is determined, if not, the previous grid is determined to be positioned on the outline of the target interest surface, and the outline of the target interest surface can be obtained by the same, in this embodiment, the grids around the center position of the target interest surface are sequentially subjected to image semantic segmentation, so that the grids positioned on the outline can be rapidly positioned, and image semantic segmentation processing is not required to be performed on all grids in the whole target area, thereby improving the efficiency of identifying the outline of the target interest surface. In addition, whether each grid around the central position of the target interest surface belongs to the target interest surface is sequentially judged through the image semantic segmentation model, specifically, whether each grid around the central position of the target interest surface belongs to a representative interest point of the central position of the target interest surface is sequentially judged through the image semantic segmentation model, and if so, the grid is determined to belong to the target interest surface. The output result of S406 is shown in fig. 6, in which the target area is the range outlined by the gray thick line, and the gray shaded portion is the range of the target interest surface.
More specifically, the image semantic segmentation model may specifically include a full convolution network (Fully Convolutional Network, FCN) and a convolution network (Convolutional Neural Networks, CNN), and further, S405 may specifically include:
for any grid, splicing a plurality of feature graphs of the grid, and then respectively inputting the feature graphs into a full convolution network of the image semantic segmentation model to respectively obtain high-dimensional features; after the high-dimensional features are spliced, inputting the spliced high-dimensional features into a convolution network of the image semantic segmentation model, and obtaining the probability that the grid belongs to the target interest surface; and if the probability is greater than a preset threshold, determining that the grid belongs to the target interest surface.
In the embodiment, considering that a single feature map is used, the features are thinner, and the amount of semantic information contained is smaller, global semantic information is fused together through the splicing of a plurality of feature maps, and the amount of semantic information of the feature map is increased, so that an image semantic segmentation model can better determine whether grids belong to a target interest plane according to the feature map, and the accuracy of image semantic segmentation is improved.
Optionally, because the feature map semantic information related to the interest point is rich, local semantic information can be represented, and the road network feature map semantic information is relatively thin, the feature map related to the interest point can be spliced into the road network feature map to obtain a global feature map, and the global feature map is input into a full convolution network, the feature map related to the interest point comprises an interest point function category feature map, an interest point name feature map, an adjacent interest point information feature map and the like, the feature map related to each interest point is respectively input into the full convolution network, and is processed in a parallel manner through a plurality of full convolution networks, each full convolution network can obtain a high-dimensional feature, and the high-dimensional feature is spliced and then is input into the convolution network of the image semantic segmentation model. It should be noted that, the feature map input by the full convolution network generally has a plurality of channels, and when a certain feature map has only one channel, the full convolution network may be replaced by a convolution network.
Optionally, as shown in fig. 7, the image semantic segmentation model of the embodiment is that feature maps related to interest points are spliced into road network feature maps to obtain global feature maps, the global feature maps are input into a full convolution network 410, interest point function class feature maps are input into a full convolution network 411, interest point name feature maps are input into the full convolution network 412, adjacent interest point information feature maps are a channel, the channel is input into the convolution network 413, high-dimensional feature output by the full convolution networks 410, 411 and 412 and the convolution network 413 are spliced and then input into the convolution network 410 for convolution processing, finally, the probability that a grid belongs to a target interest surface is obtained through an activation layer, and finally, whether the grid belongs to the target interest surface is output.
It should be noted that, the image semantic segmentation model in this embodiment may be trained in advance through training data, where the training data may be labeled feature map data, and the training process is not described herein.
According to the method, the target region which is unavailable in the satellite image can be identified based on the image semantic segmentation model by using the cheaper and more easily acquired interest point characteristic information of the target region and the road network characteristic information of the target region as the geographic attribute related data of the target region, and manual labeling based on the street view image is not needed, so that the cost can be reduced, the efficiency can be improved, and the accuracy is higher.
On the basis of the above embodiment, optionally, the geographic attribute related data of the target area further includes user behavior information in the target area, for example, comments, signs, uploading pictures, etc. of the user in the target area, and some semantic features, which are obtained after feature extraction and quantization, can represent the target area.
S403, constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid and the feature information of the road network in the grid may include:
and constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid, the feature information of the road network in the grid and the behavior information of the users in the grid.
In this embodiment, the geographic attribute related data of the target area increases the user behavior information in the target area, so that more semantic features can be added to the target area, which is convenient for further improving accuracy in the process of performing semantic segmentation on the image semantic segmentation model, and further improving accuracy in identifying the outline of the target interest surface.
On the basis of any one of the above embodiments, as shown in fig. 8, the method for identifying an interest surface may further include:
S501, dividing a target city into a plurality of areas according to road network information;
s502, receiving a target area selection instruction, and determining the target area to be identified from the areas according to the target area selection instruction.
In this embodiment, the target area to be identified is selected by the user after the plurality of areas are divided, so that the interest surface identification of one or more areas in the plurality of areas can be realized, and the user requirement can be satisfied.
Further, after the target area to be identified is determined, whether the satellite image of the target area is available or not may be determined, if the satellite image of the target area is available, S301-S304 are adopted, and if the satellite image of the target area is not available, S401-406 are adopted.
An embodiment of the present application provides an apparatus for identifying an interest surface, and fig. 9 is a block diagram of the apparatus for identifying an interest surface provided by the embodiment of the present application. As shown in fig. 9, the interest surface identifying means 600 specifically includes: an acquisition module 601, a processing module 602 and an identification module 603.
The acquiring module 601 is configured to acquire target function class information corresponding to a target interest plane in a target area to be identified, and geographic attribute related data of the target area;
The processing module 602 is configured to obtain input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area;
and the identifying module 603 is configured to identify a contour of the target interest surface in the target area according to the input data and a corresponding preset computer vision model.
In an alternative embodiment, the geographic attribute related data of the target area includes satellite images of the target area and point of interest information of the target area; the preset computer vision model is an image instance segmentation model;
the processing module 602 is configured to:
screening a set of first target interest points which are the same as the target function category in the target area according to the target function category information and the interest point information of the target area;
marking the identification of the first target interest point on the corresponding position of the satellite image of the target area according to the set of the first target interest point, and obtaining the satellite image marked with the interest point identification as input data of the image instance segmentation model;
the identifying module 603 is configured to:
And carrying out image instance segmentation on the satellite image marked with the interest point mark by the image instance segmentation model, and identifying the outline of the target interest surface in the target area.
On the basis of the above embodiment, the processing module 602 is configured to, when marking, according to the set of first target points of interest, the identification of the first target points of interest on corresponding positions of the satellite image of the target area:
and selecting the first target interest points from the first target interest point set according to a random sequence, and marking the first target interest points at the corresponding positions of the satellite images of the target area if the first target interest points are not shielded from marked interest point marks in the satellite images.
In another optional embodiment, the geographic attribute related data of the target area includes interest point feature information of the target area and road network feature information of the target area; the preset computer vision model is an image semantic segmentation model;
the processing module 602 is configured to:
determining the central position of the target interest surface according to the interest point characteristic information of the target area and the road network characteristic information of the target area;
gridding the target area, and constructing a plurality of feature graphs of each grid according to the feature information of interest points falling into the grid and the feature information of road networks in the grid;
And taking the central position of the target interest surface and the characteristic maps of each grid as input data of the image semantic segmentation model.
On the basis of the above embodiment, the identifying module 603 is configured to:
sequentially judging whether grids around the central position of the target interest surface belong to the target interest surface or not by taking the central position of the target interest surface as the center through the image semantic segmentation model;
and determining the outline of the target interest surface in the target area according to the grid positions belonging to the target interest surface.
On the basis of the above embodiment, when sequentially determining, by the image semantic segmentation model, whether each grid around the central position of the target interest surface belongs to the target interest surface, the recognition module 603 is configured to:
for any grid, splicing a plurality of feature graphs of the grid, and then respectively inputting the feature graphs into a full convolution network of the image semantic segmentation model to respectively obtain high-dimensional features;
after the high-dimensional features are spliced, inputting the spliced high-dimensional features into a convolution network of the image semantic segmentation model, and obtaining the probability that the grid belongs to the target interest surface;
and if the probability is greater than a preset threshold, determining that the grid belongs to the target interest surface.
On the basis of the above embodiment, when determining the center position of the target interest surface according to the interest point feature information of the target area and the road network feature information of the target area, the processing module 602 is configured to:
screening a second set of target interest points which are the same as the target function category from the interest points of the target area according to the target function category information;
determining a representative interest point from the second target interest point set according to the interest point characteristic information of each second target interest point in the second target interest point set, the road network characteristic information of the target area and a preset classification model;
and taking the position of the representative interest point as the central position of the target interest surface.
On the basis of the above embodiment, the feature information of the interest points in the target area includes a function class distribution of the interest points in the target area, a function class distribution of neighboring interest points of each interest point, and a location of each interest point.
On the basis of the above embodiment, the geographic attribute related data of the target area further includes user behavior information in the target area;
The processing module 602 is configured to, when constructing a plurality of feature graphs of the grid according to the feature information of the points of interest falling into the grid and the feature information of the road network within the grid:
and constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid, the feature information of the road network in the grid and the behavior information of the users in the grid.
On the basis of any of the above embodiments, the obtaining module 601 is further configured to:
dividing a target city into a plurality of areas according to road network information;
and receiving a target area selection instruction, and determining the target area to be identified from the plurality of areas according to the target area selection instruction.
The device for identifying an interest surface provided in this embodiment may be specifically used to execute the method embodiment provided in the foregoing figures, and specific functions are not provided here.
According to the interest surface identification device provided by the embodiment, the target function category information corresponding to the target interest surface in the target area to be identified and the geographic attribute related data of the target area are obtained; acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model. In the embodiment, the identification of the interest surface in the target area is realized by utilizing a computer vision technology, and the manual labeling is not needed based on the street view image, so that the cost can be reduced, the efficiency is improved, the accuracy is higher, the method and the device are applicable to different scenes, the application range is wider, and the method and the device can be applied to the identification of massive interest surfaces, so that the description of the planar information of massive interest points is possible.
According to an embodiment of the present application, the present application also provides an electronic device and a readable storage medium.
According to an embodiment of the present application, there is also provided a computer program product comprising: a computer program stored in a readable storage medium, from which at least one processor of an electronic device can read, the at least one processor executing the computer program causing the electronic device to perform the solution provided by any one of the embodiments described above.
FIG. 10 is a block diagram of an electronic device according to an embodiment of the present application. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the applications described and/or claimed herein.
As shown in fig. 10, the electronic device includes: one or more processors 701, memory 702, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions executing within the electronic device, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display device coupled to the interface. In other embodiments, multiple processors and/or multiple buses may be used, if desired, along with multiple memories and multiple memories. Also, multiple electronic devices may be connected, each providing a portion of the necessary operations (e.g., as a server array, a set of blade servers, or a multiprocessor system). One processor 701 is illustrated in fig. 10.
Memory 702 is a non-transitory computer readable storage medium provided by the present application. The memory stores instructions executable by the at least one processor to cause the at least one processor to perform the method for identifying an interest surface provided by the present application. The non-transitory computer readable storage medium of the present application stores computer instructions for causing a computer to execute the method for identifying a surface of interest provided by the present application.
The memory 702 is used as a non-transitory computer readable storage medium for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules (e.g., the acquisition module 601, the processing module 602, and the recognition module 603 shown in fig. 9) corresponding to the method for recognizing an interest surface according to an embodiment of the present application. The processor 701 executes various functional applications of the server and data processing, i.e., implements the method of identifying the surface of interest in the above-described method embodiments, by running non-transitory software programs, instructions, and modules stored in the memory 702.
Memory 702 may include a storage program area that may store an operating system, at least one application program required for functionality, and a storage data area; the storage data area may store data created according to the use of the electronic device of the interest surface recognition method, and the like. In addition, the memory 702 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 702 optionally includes memory remotely located relative to the processor 701, which may be connected to the electronic device of the face of interest recognition method via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The electronic device of the interest surface recognition method may further include: an input device 703 and an output device 704. The processor 701, the memory 702, the input device 703 and the output device 704 may be connected by a bus or otherwise, in fig. 10 by way of example.
The input device 703 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the electronic device of the face recognition method, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, a pointer stick, one or more mouse buttons, a track ball, a joystick, and the like. The output device 704 may include a display apparatus, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibration motors), among others. The display device may include, but is not limited to, a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display, and a plasma display. In some implementations, the display device may be a touch screen.
Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, application specific ASIC (application specific integrated circuit), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
These computing programs (also referred to as programs, software applications, or code) include machine instructions for a programmable processor, and may be implemented in a high-level procedural and/or object-oriented programming language, and/or in assembly/machine language. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or device (e.g., magnetic discs, optical disks, memory, programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.
The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
According to the technical scheme of the embodiment of the application, the target function category information corresponding to the target interest surface in the target area to be identified and the geographic attribute related data of the target area are obtained; acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; and identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model. In the embodiment, the identification of the interest surface in the target area is realized by utilizing a computer vision technology, and the manual labeling is not needed based on the street view image, so that the cost can be reduced, the efficiency is improved, the accuracy is higher, the method and the device are applicable to different scenes, the application range is wider, and the method and the device can be applied to the identification of massive interest surfaces, so that the description of the planar information of massive interest points is possible.
The present application also provides a computer program comprising program code which, when run by a computer, performs the method of interest surface identification as described in the above embodiments.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present application may be performed in parallel, sequentially, or in a different order, provided that the desired results of the disclosed embodiments are achieved, and are not limited herein.
The above embodiments do not limit the scope of the present application. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of the present application.

Claims (20)

1. A method of interest surface identification, comprising:
acquiring target function class information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area, wherein the geographic attribute related data of the target area comprises satellite images of the target area and interest point information of the target area;
acquiring input data of a corresponding preset computer vision model according to the target function class information and the geographic attribute related data of the target area; the preset computer vision model is an image instance segmentation model;
Identifying the outline of the target interest surface in the target area according to the input data and a corresponding preset computer vision model;
the obtaining input data of the corresponding preset computer vision model according to the target function category information and the geographic attribute related data of the target area comprises the following steps:
screening a set of first target interest points which are the same as the target function category in the target area according to the target function category information and the interest point information of the target area;
marking the identification of the first target interest point on the corresponding position of the satellite image of the target area according to the set of the first target interest point, and obtaining the satellite image marked with the interest point identification as input data of the image instance segmentation model;
the identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model comprises the following steps:
and carrying out image instance segmentation on the satellite image marked with the interest point mark by the image instance segmentation model, and identifying the outline of the target interest surface in the target area.
2. The method of claim 1, the marking the identification of the first target point of interest on the corresponding location of the satellite image of the target area according to the set of first target points of interest, comprising:
and selecting the first target interest points from the first target interest point set according to a random sequence, and marking the first target interest points at the corresponding positions of the satellite images of the target area if the first target interest points are not shielded from marked interest point marks in the satellite images.
3. The method of claim 1, wherein the geographic attribute related data of the target area includes point of interest characteristic information of the target area and road network characteristic information of the target area; the preset computer vision model is an image semantic segmentation model;
the obtaining input data of the corresponding preset computer vision model according to the target function category information and the geographic attribute related data of the target area comprises the following steps:
determining the central position of the target interest surface according to the interest point characteristic information of the target area and the road network characteristic information of the target area;
gridding the target area, and constructing a plurality of feature graphs of each grid according to the feature information of interest points falling into the grid and the feature information of road networks in the grid;
And taking the central position of the target interest surface and the characteristic maps of each grid as input data of the image semantic segmentation model.
4. The method of claim 3, wherein the identifying the outline of the target interest surface within the target area from the input data and the corresponding pre-set computer vision model comprises:
sequentially judging whether grids around the central position of the target interest surface belong to the target interest surface or not by taking the central position of the target interest surface as the center through the image semantic segmentation model;
and determining the outline of the target interest surface in the target area according to the grid positions belonging to the target interest surface.
5. The method of claim 4, wherein the sequentially determining, by the image semantic segmentation model, whether each grid around the central location of the target interest surface belongs to the target interest surface comprises:
for any grid, splicing a plurality of feature graphs of the grid, and then respectively inputting the feature graphs into a full convolution network of the image semantic segmentation model to respectively obtain high-dimensional features;
after the high-dimensional features are spliced, inputting the spliced high-dimensional features into a convolution network of the image semantic segmentation model, and obtaining the probability that the grid belongs to the target interest surface;
And if the probability is greater than a preset threshold, determining that the grid belongs to the target interest surface.
6. The method of claim 3, wherein the determining the center position of the target interest surface according to the interest point feature information of the target area and the road network feature information of the target area includes:
screening a second set of target interest points which are the same as the target function category from the interest points of the target area according to the target function category information;
determining a representative interest point from the second target interest point set according to the interest point characteristic information of each second target interest point in the second target interest point set, the road network characteristic information of the target area and a preset classification model;
and taking the position of the representative interest point as the central position of the target interest surface.
7. The method of any of claims 3-6, wherein the point of interest characteristic information of the target area includes a functional category distribution of points of interest within the target area, a functional category distribution of points of interest adjacent to each point of interest, a location where each point of interest is located.
8. The method of any of claims 3-6, wherein the geographic attribute related data of the target area further comprises user behavior information within the target area;
The constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid and the feature information of the road network in the grid comprises the following steps:
and constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid, the feature information of the road network in the grid and the behavior information of the users in the grid.
9. The method of claim 1, further comprising:
dividing a target city into a plurality of areas according to road network information;
and receiving a target area selection instruction, and determining the target area to be identified from the plurality of areas according to the target area selection instruction.
10. An apparatus for face recognition, comprising:
the acquisition module is used for acquiring target function class information corresponding to a target interest surface in a target area to be identified and geographic attribute related data of the target area, wherein the geographic attribute related data of the target area comprises satellite images of the target area and interest point information of the target area;
the processing module is used for acquiring input data of a corresponding preset computer vision model according to the target function category information and the geographic attribute related data of the target area, wherein the preset computer vision model is an image instance segmentation model;
The identification module is used for identifying the outline of the target interest surface in the target area according to the input data and the corresponding preset computer vision model;
the processing module is used for:
screening a set of first target interest points which are the same as the target function category in the target area according to the target function category information and the interest point information of the target area;
marking the identification of the first target interest point on the corresponding position of the satellite image of the target area according to the set of the first target interest point, and obtaining the satellite image marked with the interest point identification as input data of the image instance segmentation model;
the identification module is used for:
and carrying out image instance segmentation on the satellite image marked with the interest point mark by the image instance segmentation model, and identifying the outline of the target interest surface in the target area.
11. The apparatus of claim 10, the processing module, when marking the identity of the first target point of interest at a corresponding location on a satellite image of the target area from the set of first target points of interest, to:
and selecting the first target interest points from the first target interest point set according to a random sequence, and marking the first target interest points at the corresponding positions of the satellite images of the target area if the first target interest points are not shielded from marked interest point marks in the satellite images.
12. The apparatus of claim 10, wherein the geographic attribute related data of the target area comprises point of interest characteristic information of the target area and road network characteristic information of the target area; the preset computer vision model is an image semantic segmentation model;
the processing module is used for:
determining the central position of the target interest surface according to the interest point characteristic information of the target area and the road network characteristic information of the target area;
gridding the target area, and constructing a plurality of feature graphs of each grid according to the feature information of interest points falling into the grid and the feature information of road networks in the grid;
and taking the central position of the target interest surface and the characteristic maps of each grid as input data of the image semantic segmentation model.
13. The apparatus of claim 12, wherein the identification module is to:
sequentially judging whether grids around the central position of the target interest surface belong to the target interest surface or not by taking the central position of the target interest surface as the center through the image semantic segmentation model;
and determining the outline of the target interest surface in the target area according to the grid positions belonging to the target interest surface.
14. The apparatus of claim 13, wherein the means for identifying, when sequentially determining, by the image semantic segmentation model, whether each grid around a center location of the target interest surface belongs to the target interest surface, is to:
for any grid, splicing a plurality of feature graphs of the grid, and then respectively inputting the feature graphs into a full convolution network of the image semantic segmentation model to respectively obtain high-dimensional features;
after the high-dimensional features are spliced, inputting the spliced high-dimensional features into a convolution network of the image semantic segmentation model, and obtaining the probability that the grid belongs to the target interest surface;
and if the probability is greater than a preset threshold, determining that the grid belongs to the target interest surface.
15. The apparatus of claim 12, wherein the processing module, when determining the center position of the target interest surface based on the point of interest feature information of the target area and the road network feature information of the target area, is configured to:
screening a second set of target interest points which are the same as the target function category from the interest points of the target area according to the target function category information;
determining a representative interest point from the second target interest point set according to the interest point characteristic information of each second target interest point in the second target interest point set, the road network characteristic information of the target area and a preset classification model;
And taking the position of the representative interest point as the central position of the target interest surface.
16. The apparatus of any of claims 12-15, wherein the point of interest characteristic information of the target area includes a functional category distribution of points of interest within the target area, a functional category distribution of points of interest adjacent to each point of interest, a location where each point of interest is located.
17. The apparatus of any of claims 12-15, wherein the geographic attribute related data of the target area further comprises user behavior information within the target area;
the processing module is used for constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid and the feature information of the road network in the grid, wherein the feature graphs are as follows:
and constructing a plurality of feature graphs of the grid according to the feature information of the interest points falling into the grid, the feature information of the road network in the grid and the behavior information of the users in the grid.
18. The apparatus of claim 10, wherein the acquisition module is further to:
dividing a target city into a plurality of areas according to road network information;
and receiving a target area selection instruction, and determining the target area to be identified from the plurality of areas according to the target area selection instruction.
19. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein, the liquid crystal display device comprises a liquid crystal display device,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-9.
20. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-9.
CN202010516452.0A 2020-06-09 2020-06-09 Method, device, equipment and storage medium for identifying interest surface Active CN111695488B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010516452.0A CN111695488B (en) 2020-06-09 2020-06-09 Method, device, equipment and storage medium for identifying interest surface

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010516452.0A CN111695488B (en) 2020-06-09 2020-06-09 Method, device, equipment and storage medium for identifying interest surface

Publications (2)

Publication Number Publication Date
CN111695488A CN111695488A (en) 2020-09-22
CN111695488B true CN111695488B (en) 2023-09-08

Family

ID=72479842

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010516452.0A Active CN111695488B (en) 2020-06-09 2020-06-09 Method, device, equipment and storage medium for identifying interest surface

Country Status (1)

Country Link
CN (1) CN111695488B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112016326A (en) * 2020-09-25 2020-12-01 北京百度网讯科技有限公司 Map area word recognition method and device, electronic equipment and storage medium
CN112541479B (en) * 2020-12-25 2024-01-05 北京百度网讯科技有限公司 Panorama and interest point hooking method and device, electronic equipment and storage medium
CN112784091A (en) * 2021-01-26 2021-05-11 上海商汤科技开发有限公司 Interest analysis method and related device and equipment
CN113536132A (en) * 2021-07-28 2021-10-22 拉扎斯网络科技（上海）有限公司 Interest plane AOI processing method and device, electronic equipment and storage medium
CN113377783B (en) * 2021-08-12 2021-12-10 腾讯科技（深圳）有限公司 Data processing method and device, electronic equipment and computer readable storage medium
CN113723405A (en) * 2021-08-31 2021-11-30 北京百度网讯科技有限公司 Method and device for determining area outline and electronic equipment
CN113793346B (en) * 2021-09-16 2022-08-30 北京百度网讯科技有限公司 Region attribute method and device, electronic equipment and readable storage medium
CN114322985B (en) * 2021-12-24 2024-04-12 深圳依时货拉拉科技有限公司 Electronic map recommendation point display method, device, equipment and storage medium
CN116863127A (en) * 2022-03-28 2023-10-10 华为技术有限公司 Method for acquiring region of interest and related equipment
CN115641430B (en) * 2022-10-12 2024-01-30 阿里巴巴（中国）有限公司 Method, device, medium and computer equipment for determining interest surface
CN116843892B (en) * 2023-08-02 2024-04-26 深圳市名通科技股份有限公司 AOI scene contour recognition method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109359170A (en) * 2018-11-02 2019-02-19 百度在线网络技术（北京）有限公司 Method and apparatus for generating information
CN109685060A (en) * 2018-11-09 2019-04-26 科大讯飞股份有限公司 Image processing method and device
CN109710709A (en) * 2018-12-13 2019-05-03 北京百度网讯科技有限公司 Interest point data processing method, device, electronic equipment and storage medium
CN110726418A (en) * 2019-10-10 2020-01-24 北京百度网讯科技有限公司 Method, device and equipment for determining interest point region and storage medium

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10609284B2 (en) * 2016-10-22 2020-03-31 Microsoft Technology Licensing, Llc Controlling generation of hyperlapse from wide-angled, panoramic videos

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109359170A (en) * 2018-11-02 2019-02-19 百度在线网络技术（北京）有限公司 Method and apparatus for generating information
CN109685060A (en) * 2018-11-09 2019-04-26 科大讯飞股份有限公司 Image processing method and device
CN109710709A (en) * 2018-12-13 2019-05-03 北京百度网讯科技有限公司 Interest point data processing method, device, electronic equipment and storage medium
CN110726418A (en) * 2019-10-10 2020-01-24 北京百度网讯科技有限公司 Method, device and equipment for determining interest point region and storage medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陈凯 ; 宋晓 ; 刘敬 ; .基于深度卷积网络与尺度不变特征变换的行人跟踪框架.中国科学:信息科学.2018,(第07期),全文. *

Also Published As

Publication number Publication date
CN111695488A (en) 2020-09-22

Similar Documents

Publication Publication Date Title
CN111695488B (en) Method, device, equipment and storage medium for identifying interest surface
CN110726418B (en) Method, device and equipment for determining interest point region and storage medium
KR102652742B1 (en) Method and device for identifying updated road, electronic equipment and computer storage medium
CN111797187B (en) Map data updating method and device, electronic equipment and storage medium
US11361005B2 (en) Method for processing map data, device, and storage medium
CN110675644B (en) Method and device for identifying road traffic lights, electronic equipment and storage medium
CN112528786B (en) Vehicle tracking method and device and electronic equipment
CN111860304A (en) Image labeling method, electronic device, equipment and storage medium
CN112013865B (en) Method, system, electronic device and medium for determining traffic gate
CN111291681B (en) Method, device and equipment for detecting lane change information
CN113091757B (en) Map generation method and device
CN111597986B (en) Method, apparatus, device and storage medium for generating information
CN112581533B (en) Positioning method, positioning device, electronic equipment and storage medium
CN113160693B (en) Road intersection processing method, device, equipment and storage medium
CN114443794A (en) Data processing and map updating method, device, equipment and storage medium
CN111597987B (en) Method, apparatus, device and storage medium for generating information
CN111523061B (en) Method and apparatus for generating interest plane
CN114037966A (en) High-precision map feature extraction method, device, medium and electronic equipment
CN114186007A (en) High-precision map generation method and device, electronic equipment and storage medium
CN113408662A (en) Image recognition method and device, and training method and device of image recognition model
CN113011298A (en) Truncated object sample generation method, target detection method, road side equipment and cloud control platform
CN113297878B (en) Road intersection identification method, device, computer equipment and storage medium
CN112990241B (en) Track matching method, device, equipment and storage medium
CN111324616A (en) Method, device and equipment for detecting lane line change information
CN113256642B (en) Method, device, equipment and storage medium for determining failure area in electronic map

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
GR01 Patent grant
GR01 Patent grant