Urban space entity recognition method and device

Abstract

The present disclosure provides a method and device for identifying urban space entities. The urban space entity identification device divides an aerial image according to a predetermined size to obtain a plurality of raster images; for each raster image, a POI category corresponding to the location information of the raster image is queried from a preset POI list; a spatial entity training set associated with the POI category is queried from a spatial model set; a machine learning model is trained using the spatial entity training set; and the raster image is identified using the trained machine learning model to obtain the type of the urban space entity in the raster image. The present disclosure can effectively improve the recognition efficiency of urban space entities and effectively reduce the computing overhead of the machine.

Description

Technical Field

[0001] The present disclosure relates to the field of information processing, and in particular to a method and device for identifying urban space entities. Background Art

[0002] Identifying urban space entities to distinguish different physical objects in the city is one of the supporting technologies for digital twin city business.

[0003] At present, in the recognition of urban space entities, machine learning models (for example, CNN (Convolutional Neural Network)) are usually used for image recognition to identify the type of urban space entities. Summary of the invention

[0004] The inventors found through research that in the process of using machine learning models for image recognition, this requires the establishment of a huge urban graphic sample library for manual annotation. On the one hand, the annotation is time-consuming and labor-intensive, and the cost is extremely high. On the other hand, when judging the image features, the entire sample library needs to be matched, which is inefficient.

[0005] Based on this, the present disclosure provides a solution that can realize urban space entity recognition conveniently and quickly.

[0006] According to a first aspect of an embodiment of the present disclosure, a method for identifying urban spatial entities is provided, comprising: segmenting an aerial image according to a predetermined size to obtain a plurality of raster images; for each raster image, querying a POI category corresponding to the location information of the raster image from a preset point of interest (POI) list; querying a spatial entity training set associated with the POI category from a spatial model set; training a machine learning model using the spatial entity training set; and identifying the raster image using the trained machine learning model to obtain the type of the urban spatial entity in the raster image.

[0007] In some embodiments, the above method further includes: performing semantic recognition on the POI information in the original POI list to obtain the POI category of the POI information; and inserting the POI category into the original POI list to generate the preset POI list.

[0008] In some embodiments, the position information of the raster image is the position information of the geometric center of the raster image.

[0009] In some embodiments, the above method further includes: storing the type of urban space entity in the raster image.

[0010] According to a second aspect of an embodiment of the present disclosure, there is provided an urban space entity recognition device, comprising: an image segmentation module, configured to segment an aerial image according to a predetermined size to obtain a plurality of raster images; a POI category query module, configured to query, for each raster image, a POI category corresponding to the location information of the raster image from a preset point of interest (POI) list; a training set query module, configured to query a spatial entity training set associated with the POI category from a spatial model set; a training module, configured to train a machine learning model using the spatial entity training set; and an entity recognition module, configured to recognize the raster image using the trained machine learning model to obtain the type of the urban space entity in the raster image.

[0011] In some embodiments, the above-mentioned device also includes: a list generation module, which is configured to perform semantic recognition on the POI information in the original POI list to obtain the POI category of the POI information, and insert the POI category into the original POI list to generate the preset POI list.

[0012] In some embodiments, the position information of the raster image is the position information of the geometric center of the raster image.

[0013] In some embodiments, the above-mentioned apparatus further includes: a storage module configured to store the types of urban space entities in the raster image.

[0014] According to a third aspect of an embodiment of the present disclosure, there is provided an urban space entity recognition device, comprising: a memory configured to store instructions; a processor coupled to the memory, the processor being configured to execute a method as described in any of the above embodiments based on the instructions stored in the memory.

[0015] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method involved in any of the above embodiments is implemented.

[0016] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0018] The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0019] Figure 1is a flow chart of a method for identifying urban space entities according to an embodiment of the present disclosure;

[0020] Figure 2 is a schematic diagram of the structure of an urban space entity recognition device according to an embodiment of the present disclosure;

[0021] Figure 3 is a schematic structural diagram of an urban space entity recognition device according to another embodiment of the present disclosure;

[0022] Figure 4 It is a structural schematic diagram of an urban space entity recognition device according to another embodiment of the present disclosure.

[0023] It should be understood that the size of each part shown in the accompanying drawings is not drawn according to the actual proportional relationship. In addition, the same or similar reference numerals represent the same or similar components. DETAILED DESCRIPTION

[0024] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and is in no way intended to limit the present disclosure and its application or use. The present disclosure can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the present disclosure thorough and complete and to fully express the scope of the present disclosure to those skilled in the art. It should be noted that unless otherwise specifically stated, the relative arrangement of the components and steps, the composition of the materials, and the numerical values ​​set forth in these embodiments should be interpreted as being merely exemplary and not as limiting.

[0025] The words “include” or “comprising” and the like used in the present disclosure mean that the elements preceding the words include the elements listed after the words, and do not exclude the possibility of also including other elements.

[0026] All terms (including technical terms or scientific terms) used in the present disclosure have the same meanings as those understood by ordinary technicians in the field to which the present disclosure belongs, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries, such as general dictionaries, should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined herein.

[0027] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0028] Figure 1 1 is a flow chart of a method for identifying urban space entities according to an embodiment of the present disclosure. In some embodiments, the following method steps are performed by an urban space entity identification device.

[0029] In step 101, the aerial image is segmented according to a predetermined size to obtain a plurality of raster images.

[0030] For example, drones are used to take photos over a city to obtain aerial images.

[0031] In some embodiments, a conditional random field model is applied for image segmentation.

[0032] In step 102, for each raster image, a POI category corresponding to the location information of the raster image is searched from a preset POI (Point of Interest) list.

[0033] In some embodiments, the position information of the raster image is the position information of the geometric center of the raster image. For example, the position information of the raster image is the longitude and latitude coordinates of the geometric center of the raster image.

[0034] In some embodiments, semantic recognition is performed on the POI information in the original POI list to obtain the POI category of the POI information, and the POI category is inserted into the original POI list to generate a preset POI list.

[0035] In step 103, a spatial entity training set associated with the POI category is queried from the spatial model set.

[0036] In step 104, the machine learning model is trained using the spatial entity training set.

[0037] In step 105, the raster image is identified using the trained machine learning model to obtain the type of the urban space entity in the raster image.

[0038] In some embodiments, the types of urban spatial entities in the raster image are stored so that the identified urban spatial entity types can be used for digital twin city services.

[0039] For example, the original POI list is shown in Table 1.

[0040] POI Latitude and longitude China Telecom Beijing Technology Innovation Park 40.1122863,116.4764840 Tongtai Building 39.9164351,116.3647601 Financial Street Plaza 39.9158961,116.3638427 …… ……

[0041] Table 1

[0042] Next, semantic recognition is performed on the POI information in Table 1 to obtain the POI category of the POI information, and the POI category is inserted into the original POI list to generate a preset POI list, as shown in Table 2.

[0043] POI Latitude and longitude category China Telecom Beijing Technology Innovation Park 40.1122863,116.4764840 Park Tongtai Building 39.9164351,116.3647601 Buildings Financial Street Plaza 39.9158961,116.3638427 square …… …… ……

[0044] Table 2

[0045] In addition, in the pre-labeled spatial model set, graphics associated with each POI category are set as spatial entity training sets, as shown in Table 3.

[0046]

[0047]

[0048] Table 3

[0049] Next, the aerial image is segmented according to a predetermined size to obtain multiple raster images. According to the latitude and longitude information of a raster image, the area is queried from the preset POI list as Tongtai Building, and the corresponding POI category is building. Then, the spatial entity training set associated with the building is queried from the spatial model set as Figures 4, 5, and 6. By using Figures 4, 5, and 6 as training samples, a machine learning model such as CNN is trained. Finally, the trained machine learning model is used to identify the raster image to obtain the type of urban spatial entity in the raster image. The final results are shown in Table 4.

[0050] Figure Number Graphics Category Graphic Description Graphics picture folder 7 Roof Bungalow roof A 8 Roof Gable roof B 9 Roof Hexagonal roof C 10 Roof Tower roof D …… …… …… ……

[0051] Table 4

[0052] In the process of training the machine learning model, since only a limited sample set is used, that is, graphics associated with the POI category are used as samples, and the trained machine learning model only recognizes images with the POI category, it can effectively improve the recognition efficiency and effectively reduce the machine's computing overhead.

[0053] Figure 2 It is a structural schematic diagram of an urban space entity recognition device according to an embodiment of the present disclosure.

[0054] like Figure 2 As shown, the urban space entity recognition device includes: an image segmentation module 21, a POI category query module 22, a training set query module 23, a training module 24 and an entity recognition module 25.

[0055] The image segmentation module 21 is configured to segment the aerial image according to a predetermined size to obtain a plurality of raster images.

[0056] For example, drones are used to take photos over a city to obtain aerial images.

[0057] In some embodiments, a conditional random field model is applied for image segmentation.

[0058] The POI category query module 22 is configured to query, for each raster image, a POI category corresponding to the location information of the raster image from a preset POI list.

[0059] In some embodiments, the position information of the raster image is the position information of the geometric center of the raster image. For example, the position information of the raster image is the longitude and latitude coordinates of the geometric center of the raster image.

[0060] The training set query module 23 is configured to query the spatial entity training set associated with the POI category from the spatial model set.

[0061] The training module 24 is configured to train the machine learning model using the spatial entity training set.

[0062] The entity recognition module 25 is configured to use a trained machine learning model to recognize the raster image to obtain the type of the urban space entity in the raster image.

[0063] Figure 3 It is a structural schematic diagram of an urban space entity recognition device according to another embodiment of the present disclosure.

[0064] Figure 3 and Figure 2 The difference is that in Figure 3 In the illustrated embodiment, the urban space entity recognition device further includes a list generation module 26 .

[0065] The list generation module 26 is configured to perform semantic recognition on the POI information in the original POI list to obtain the POI category of the POI information, and insert the POI category into the original POI list to generate a preset POI list.

[0066] In some embodiments, Figure 3 As shown, the urban space entity recognition device also includes a storage module 27.

[0067] The storage module 27 is configured to store the types of urban space entities in the raster image so that the identified urban space entity types can be used for digital twin city services.

[0068] Figure 4 FIG. 1 is a schematic diagram of the structure of an urban space entity recognition device according to another embodiment of the present disclosure. Figure 4 As shown, the device includes a memory 41 and a processor 42.

[0069] The memory 41 is used to store instructions. The processor 42 is coupled to the memory 41. The processor 42 is configured to execute the instructions stored in the memory to implement the following Figure 1 The method of any one of the embodiments.

[0070] like Figure 4 As shown, the device also includes a communication interface 43 for information exchange with other devices. At the same time, the device also includes a bus 44, through which the processor 42, the communication interface 43, and the memory 41 communicate with each other.

[0071] The memory 41 may include high-speed RAM (Random Access Memory) and may also include NVM (Non-Volatile Memory). For example, at least one disk storage. The memory 41 may also be a storage array. The memory 41 may also be divided into blocks, and the blocks may be combined into virtual volumes according to certain rules.

[0072] In addition, the processor 42 may be a central processing unit, or may be an ASIC (Application Specific Integrated Circuit), or may be configured to implement one or more integrated circuits of the embodiments of the present disclosure.

[0073] The present disclosure also provides a computer-readable storage medium. The computer-readable storage medium stores computer instructions, which are executed by a processor to implement the following Figure 1 The method of any one of the embodiments.

[0074] In some embodiments, the above-mentioned functional modules can be implemented as a general-purpose processor, a programmable logic controller (PLC), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components or any appropriate combination thereof for performing the functions described in the present disclosure.

[0075] So far, the embodiments of the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solutions disclosed herein.

[0076] Although some specific embodiments of the present disclosure have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified or some technical features may be replaced by equivalents without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A method for identifying urban space entities, comprising: Segmenting the aerial image according to a predetermined size to obtain a plurality of raster images; For each raster image, querying a preset point of interest (POI) list for a POI category corresponding to the location information of the raster image; Querying a spatial entity training set associated with the POI category from the spatial model set; Training a machine learning model using the spatial entity training set; The raster image is identified using a trained machine learning model to obtain the type of urban space entity in the raster image.

2. The method according to claim 1, further comprising: Performing semantic recognition on the POI information in the original POI list to obtain the POI category of the POI information; The POI category is inserted into the original POI list to generate the preset POI list.

3. The method according to claim 1, wherein: The position information of the raster image is the position information of the geometric center of the raster image.

4. The method according to any one of claims 1 to 3, further comprising: The types of urban space entities in the raster image are stored.

5. An urban space entity recognition device, comprising: An image segmentation module is configured to segment the aerial image according to a predetermined size to obtain a plurality of raster images; A POI category query module is configured to query, for each raster image, a POI category corresponding to the location information of the raster image from a preset point of interest POI list; A training set query module is configured to query a spatial entity training set associated with the POI category from a spatial model set; A training module, configured to train a machine learning model using the spatial entity training set; The entity recognition module is configured to use a trained machine learning model to recognize the raster image to obtain the type of urban space entity in the raster image.

6. The apparatus according to claim 5, further comprising: The list generation module is configured to perform semantic recognition on the POI information in the original POI list to obtain the POI category of the POI information, and insert the POI category into the original POI list to generate the preset POI list.

7. The device according to claim 5, wherein: The position information of the raster image is the position information of the geometric center of the raster image.

8. The device according to any one of claims 5 to 7, further comprising: The storage module is configured to store the types of urban space entities in the raster image.

9. An urban space entity recognition device, comprising: a memory configured to store instructions; A processor is coupled to the memory, and the processor is configured to execute the method according to any one of claims 1 to 4 based on instructions stored in the memory.

10. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images