A method and system for collecting building information data

Abstract

The present invention provides a method and system for collecting building information data, which relate to the field of data processing technology. In the present invention, the original image set is split to form multiple original image subsets; each original image subset is marked as an image sequence to be analyzed, and the shallow key information of the building corresponding to the image sequence to be analyzed is extracted; the deep key information of the building corresponding to the image sequence to be analyzed is mined through the target image analysis neural network; the image abnormality analysis data corresponding to the image sequence to be analyzed is analyzed through the target image analysis neural network based on the shallow key information of the building and the deep key information of the building; based on the image abnormality analysis data, multiple original image subsets are screened to collect the target building image in the original image set. Based on the above method, the reliability of building data collection can be improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for collecting building information data. Background Art

[0002] After monitoring a building, such as performing image monitoring, it is generally necessary to screen and collect target monitoring images from the collected monitoring images. For example, the target monitoring images can be screened and collected based on the analysis results of the monitoring images, such as the abnormality analysis results. However, in the prior art, the abnormality analysis is generally performed manually, which results in low reliability of the image abnormality analysis, and thus leads to low reliability of the screening collection based on the abnormality analysis results. Summary of the invention

[0003] In view of this, an object of the present invention is to provide a method and system for collecting building information data, so as to improve the reliability of building data collection to a certain extent.

[0004] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0005] A method for collecting building information data, comprising:

[0006] Extracting an original image set corresponding to the building to be analyzed, and splitting the original image set to form a plurality of original image subsets corresponding to the original image set;

[0007] Marking each of the original image subsets as an image sequence to be analyzed, and extracting shallow key information of the building corresponding to the image sequence to be analyzed, wherein the image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed;

[0008] Mining out the deep key information of the building corresponding to the image sequence to be analyzed through the target image analysis neural network;

[0009] By means of the target image analysis neural network, according to the shallow key information of the building and the deep key information of the building, image anomaly analysis data corresponding to the image sequence to be analyzed is analyzed, and the image anomaly analysis data is used to reflect whether there are abnormal images or the degree of image anomaly of the image of the building to be analyzed included in the image sequence to be analyzed;

[0010] Based on the image anomaly analysis data, the plurality of original image subsets are screened to capture a target building image in the original image set.

[0011] In some preferred embodiments, in the above-mentioned method for collecting building information data, the method for collecting building information data further includes:

[0012] An exemplary building image sequence is extracted, wherein the number of image frames of the exemplary building image sequence does not exceed the number of preconfigured reference image frames, the exemplary building image sequence has image anomaly annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of the exemplary building image;

[0013] According to the target sequence length, performing sliding window segmentation processing on the first number of building image subsequences to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences;

[0014] According to the exemplary building image sequence, the initial image analysis neural network is initially optimized to form an intermediate image analysis neural network;

[0015] performing image anomaly analysis on the plurality of exemplary image subsequences by means of the intermediate image analysis neural network, and determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences;

[0016] Based on the exemplary building image sequence and the first exemplary image subsequence, the intermediate image analysis neural network is subjected to network optimization processing to form a target image analysis neural network corresponding to the intermediate image analysis neural network.

[0017] In some preferred embodiments, in the above-mentioned method for collecting building information data, the method for collecting building information data further includes:

[0018] Performing inter-frame comparison analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, an image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity;

[0019] The exemplary building image sequence is segmented according to the inter-frame difference identification information to form the first number of building image sub-sequences.

[0020] In some preferred embodiments, in the above-mentioned method for collecting building information data, each exemplary image subsequence in the multiple exemplary image subsequences is sequentially marked as a to-be-processed image subsequence, the to-be-processed image subsequence has image anomaly annotation data, and the image anomaly assessment data of the to-be-processed image subsequence analyzed by the intermediate image analysis neural network includes image anomaly characterization data corresponding to the to-be-processed image subsequence and an assessment possibility parameter for the image anomaly characterization data, and the image anomaly characterization data is used to reflect whether there is an abnormal image or the degree of image anomaly;

[0021] The step of determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences comprises:

[0022] When the image anomaly characterization data corresponding to the image subsequence to be processed and the image anomaly annotation data corresponding to the image subsequence to be processed are consistent, and when the evaluation possibility parameter for the image subsequence to be processed is not less than a pre-configured reference evaluation possibility parameter, the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked as a first exemplary image subsequence.

[0023] In some preferred embodiments, in the above-mentioned method for collecting building information data, each exemplary image subsequence in the multiple exemplary image subsequences is sequentially marked as a to-be-processed image subsequence, the to-be-processed image subsequence has image anomaly annotation data, and the image anomaly assessment data corresponding to the to-be-processed image subsequence analyzed by the intermediate image analysis neural network includes image anomaly characterization data corresponding to the to-be-processed image subsequence, and the image anomaly characterization data is used to reflect whether there is an abnormal image or the degree of image anomaly;

[0024] The step of determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences comprises:

[0025] In the case where the image anomaly characterization data corresponding to the image subsequence to be processed is consistent with the image anomaly annotation data corresponding to the image subsequence to be processed, the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked to form a first exemplary image subsequence.

[0026] In some preferred embodiments, in the above-mentioned method for collecting building information data, each image sequence or image subsequence in the exemplary building image sequence and the first exemplary image subsequence is sequentially marked as data for network optimization;

[0027] The step of performing network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network includes:

[0028] Extracting the shallow key information of the exemplary building contained in the network optimization data;

[0029] Through the intermediate image analysis neural network, deep key information of the exemplary building corresponding to the network optimization data is mined;

[0030] By means of the intermediate image analysis neural network, based on the shallow key information of the exemplary building and the deep key information of the exemplary building, image anomaly assessment data corresponding to the network optimization data is analyzed;

[0031] Analyze the image anomaly analysis cost value corresponding to the intermediate image analysis neural network based on the image anomaly assessment data corresponding to the network optimization data and the image anomaly labeling data of the network optimization data;

[0032] According to the image anomaly analysis cost value, the intermediate image analysis neural network is subjected to network optimization processing to form a corresponding target image analysis neural network.

[0033] In some preferred embodiments, in the above-mentioned method for collecting building information data, the exemplary shallow key information of the building includes:

[0034] Key information of building components corresponding to the building components of the exemplary building images in the network optimization data, the key information of building components being formed based on mining the type ratio information of the building components of each exemplary building image in the network optimization data, and the expression form of the key information of building components including vectors;

[0035] image cumulative key information of a subsequence of building images in the network optimization data, the image cumulative key information being formed based on mining the number of image frames and the amount of image data of exemplary building images included in the subsequence of building images, and the expression form of the image cumulative key information comprising a vector; and / or

[0036] The component cumulative key information of the main building components in the building components of the exemplary building images in the network optimization data is formed based on mining the component quantity and component type quantity of the building components in each exemplary building image in the network optimization data, and the expression form of the component cumulative key information includes a vector.

[0037] The embodiment of the present invention further provides a system for collecting building information data, including:

[0038] An original image processing module is used to extract an original image set corresponding to the building to be analyzed, and split the original image set to form a plurality of original image subsets corresponding to the original image set;

[0039] A key information extraction module is used to mark each of the original image subsets as an image sequence to be analyzed, and extract shallow key information of the building corresponding to the image sequence to be analyzed, wherein the image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed;

[0040] A key information mining module is used to mine the deep key information of the building corresponding to the image sequence to be analyzed through a target image analysis neural network;

[0041] An abnormality analysis module is used to analyze the image abnormality analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network, and the image abnormality analysis data is used to reflect whether there is an abnormal image or the degree of image abnormality in the image of the building to be analyzed included in the image sequence to be analyzed;

[0042] The image acquisition module is used to screen the multiple original image subsets based on the image anomaly analysis data to acquire the target building image from the original image set.

[0043] In some preferred embodiments, in the above-mentioned building information data collection system, the building information data collection system further includes:

[0044] An exemplary image extraction module is used to extract an exemplary building image sequence, wherein the number of image frames of the exemplary building image sequence does not exceed the number of pre-configured reference image frames, the exemplary building image sequence has image anomaly annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of the exemplary building image;

[0045] a sliding window segmentation processing module, configured to perform sliding window segmentation processing on the first number of building image subsequences according to a target sequence length, so as to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences;

[0046] An initial optimization processing module, used for performing initial optimization processing on the initial image analysis neural network according to the exemplary building image sequence to form an intermediate image analysis neural network;

[0047] a subsequence determination module, configured to perform image anomaly analysis on the plurality of exemplary image subsequences through the intermediate image analysis neural network, and determine a first exemplary image subsequence from the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences;

[0048] A network optimization processing module is used to perform network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network.

[0049] In some preferred embodiments, in the above-mentioned building information data collection system, the building information data collection system further includes:

[0050] An inter-frame comparison and analysis module is used to perform inter-frame comparison and analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, an image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity;

[0051] The segmentation processing module is used to perform segmentation processing on the exemplary building image sequence according to the inter-frame difference identification information to form the first number of building image sub-sequences.

[0052] The embodiment of the present invention provides a method and system for collecting building information data, which can split the original image set to form multiple original image subsets; mark each original image subset as an image sequence to be analyzed, and extract the shallow key information of the building corresponding to the image sequence to be analyzed; mine the deep key information of the building corresponding to the image sequence to be analyzed through the target image analysis neural network; analyze the image abnormality analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network; based on the image abnormality analysis data, screen multiple original image subsets to collect the target building image in the original image set. Based on the above steps, since it no longer relies on the low-precision manual abnormality analysis, but adopts the neural network for analysis, and the basis of the neural network analysis includes the shallow key information of the building and the deep key information of the building, the analysis basis is more sufficient, therefore, the reliability of building data collection can be improved to a certain extent, thereby improving the problem of low reliability of building data collection existing in the prior art.

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a structural block diagram of a building information data collection platform provided in an embodiment of the present invention.

[0055] Figure 2 A flowchart of the steps of the method for collecting building information data provided by an embodiment of the present invention.

[0056] Figure 3 A schematic diagram of various modules included in the building information data collection system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0058] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] like Figure 1 As shown, an embodiment of the present invention provides a building information data collection platform, wherein the building information data collection platform may include a memory and a processor.

[0060] In detail, the memory and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, they can be electrically connected to each other through one or more communication buses or signal lines. The memory can store at least one software function module (computer program) that can exist in the form of software or firmware. The processor can be used to execute the executable computer program stored in the memory, thereby realizing the method for collecting building information data provided by the embodiment of the present invention (as described later).

[0061] Specifically, in one embodiment, the memory may be, but not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc. The processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be 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.

[0062] Specifically, in one implementation, the building information data collection platform may be a server with data processing capabilities.

[0063] Combination Figure 2The embodiment of the present invention also provides a method for collecting building information data, which can be applied to the above-mentioned building information data collection platform. Among them, the method steps defined in the process related to the building information data collection method can be implemented by the building information data collection platform. Figure 2 The specific process shown is explained in detail.

[0064] Step S110 , extracting an original image set corresponding to the building to be analyzed, and splitting the original image set to form a plurality of original image subsets corresponding to the original image set.

[0065] In an embodiment of the present invention, the building information data acquisition platform can extract the original image set corresponding to the building to be analyzed (for example, after the image acquisition terminal device acquires the original image set, it can send it to the building information data acquisition platform), and split the original image set to form a plurality of original image subsets corresponding to the original image set (exemplarily, the original image set and each of the original image subsets belong to an ordered set, that is, the images of the building to be analyzed included are sorted in chronological order according to the acquisition time).

[0066] Step S120 , marking each of the original image subsets as an image sequence to be analyzed, and extracting shallow key information of the building corresponding to the image sequence to be analyzed.

[0067] In an embodiment of the present invention, the acquisition platform of the building information data can mark each of the original image subsets as an image sequence to be analyzed (to perform subsequent processing respectively), and extract the shallow key information of the building corresponding to the image sequence to be analyzed. The image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed.

[0068] Step S130, mining the deep key information of the building corresponding to the image sequence to be analyzed through the target image analysis neural network.

[0069] In an embodiment of the present invention, the building information data collection platform can mine the deep key information of the building corresponding to the image sequence to be analyzed through a target image analysis neural network (the target image analysis neural network can be formed through network optimization processing so that it has a more reliable ability to mine the deep key information of the building).

[0070] Step S140, analyzing the image anomaly analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network.

[0071] In an embodiment of the present invention, the acquisition platform of the building information data can analyze the image anomaly analysis data corresponding to the image sequence to be analyzed according to the shallow key information of the building and the deep key information of the building through the target image analysis neural network (exemplarily, since the dimensions represented by the shallow key information of the building and the deep key information of the building are different, at least the determination method is different, therefore, in order to fully combine the two aspects of key information in the process of analyzing the image anomaly analysis data, the shallow key information of the building and the deep key information of the building can be spliced ​​to form a spliced ​​key information, and then the spliced ​​key information can be analyzed. Specifically, an excitation mapping output can be performed, such as implemented by a softmax function, to output the corresponding image anomaly analysis data). The image anomaly analysis data is used to reflect whether there are abnormal images or the degree of image anomaly in the image of the building to be analyzed included in the image sequence to be analyzed (specifically, it can be configured according to actual needs).

[0072] Step S150: screening the plurality of original image subsets based on the image anomaly analysis data to capture a target building image in the original image set.

[0073] In an embodiment of the present invention, the building information data collection platform can screen the multiple original image subsets based on the image anomaly analysis data to collect the target building image in the original image set (exemplarily, each original image subset with anomalies or an image anomaly degree greater than a pre-configured reference image anomaly degree can be screened out, or some images in each original image subset with anomalies or an image anomaly degree greater than a pre-configured reference image anomaly degree can be screened out, so that all images that have not been screened out can be marked as target building images. In addition, the specific definition of image anomaly is not limited and can be configured according to specific application scenarios, such as it can refer to inconsistencies in the content of the image, or it can also refer to untrue content in the image, that is, the image has been tampered with. Problem).

[0074] Based on the aforementioned steps, such as the aforementioned steps S110 to S150, since it no longer relies on the inaccurate manual abnormality analysis, but adopts the neural network analysis, and the basis of the neural network analysis includes the shallow key information of the building and the deep key information of the building, the analysis basis is more sufficient. Therefore, the reliability of building data collection can be improved to a certain extent, thereby improving the problem of low reliability of building data collection existing in the prior art.

[0075] Specifically, in one implementation, the method for collecting building information data may further include the following specific contents:

[0076] An exemplary building image sequence is extracted, the number of image frames of the exemplary building image sequence does not exceed the pre-configured number of reference image frames (the specific value of the reference image frame number is not limited and can be configured according to actual needs. In this way, the data demand for network optimization processing can be reduced to a certain extent through the configuration of the reference image frame number), the exemplary building image sequence has image abnormality annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of exemplary building image (that is, it can include one frame of exemplary building image, and can also include multiple frames of exemplary building images, and the multiple frames of exemplary building images can be adjacent, that is, the corresponding acquisition time is continuous);

[0077] According to a target sequence length (the specific value of the target sequence length is not limited, such as 1, 2, 3, 4, etc.), the first number of building image subsequences are subjected to sliding window segmentation processing to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences (e.g., the number of building image subsequences included in each of the exemplary image subsequences is equal to the target sequence length);

[0078] According to the exemplary building image sequence, the initial image analysis neural network is initially optimized to form an intermediate image analysis neural network (exemplarily, exemplary key information corresponding to the exemplary building image sequence can be analyzed according to the building image subsequence included in the exemplary building image sequence, and image anomaly assessment data corresponding to the exemplary building image sequence, as well as image anomaly assessment data corresponding to the exemplary building image sequence and image anomaly annotation data corresponding to the exemplary building image sequence are analyzed by the initial image analysis neural network according to the exemplary key information, and the corresponding image anomaly analysis cost value, i.e., error, is analyzed and output, and the initial image analysis neural network is subjected to network optimization processing according to the image anomaly analysis cost value to form an intermediate image analysis neural network);

[0079] performing image anomaly analysis on the plurality of exemplary image subsequences by means of the intermediate image analysis neural network, and determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences;

[0080] Based on the exemplary building image sequence and the first exemplary image subsequence, the intermediate image analysis neural network is subjected to network optimization processing (such as the aforementioned initial optimization processing) to form a target image analysis neural network corresponding to the intermediate image analysis neural network (based on this, by determining the first exemplary image subsequence with high precision, reliable network optimization processing of the intermediate image analysis neural network can be achieved, thereby improving the analysis capability of the formed target image analysis neural network to a certain extent and ensuring its analysis accuracy).

[0081] Specifically, in one implementation, the method for collecting building information data may further include the following specific contents:

[0082] Performing inter-frame comparison analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, the image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity (that is, the image similarity may be calculated for each two adjacent frames of exemplary building images in the exemplary building image sequence, and then the image similarity is compared with the reference image similarity, wherein the image similarity may be calculated by calculating the overlap according to the distribution of the extracted key points, or by calculating the image similarity of the two frames of exemplary building images based on a corresponding neural network);

[0083] Based on the inter-frame difference identification information, the exemplary building image sequence is segmented to form the first number of building image sub-sequences (exemplarily, the two frames of exemplary building images corresponding to the sequence positions corresponding to the inter-frame difference identification information can be allocated to two adjacent building image sub-sequences, respectively as the last data and the first data of the sub-sequences).

[0084] Specifically, in one embodiment, each of the multiple exemplary image subsequences is marked as a to-be-processed image subsequence in turn, and the to-be-processed image subsequence has image anomaly annotation data. The image anomaly assessment data of the to-be-processed image subsequence analyzed by the intermediate image analysis neural network includes image anomaly characterization data corresponding to the to-be-processed image subsequence and an assessment possibility parameter (i.e., the possibility of anomaly, or the possibility of the degree of anomaly) for the image anomaly characterization data. The image anomaly characterization data is used to reflect whether there is an abnormal image or the degree of image anomaly. Based on this, the step of determining a first exemplary image subsequence among the multiple exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the multiple exemplary image subsequences may further include the following specific contents:

[0085] When the image anomaly characterization data corresponding to the image subsequence to be processed is consistent with the image anomaly annotation data corresponding to the image subsequence to be processed, and when the evaluation possibility parameter for the image subsequence to be processed is not less than a pre-configured reference evaluation possibility parameter (the specific data of the reference evaluation possibility parameter is not limited and can be configured according to actual needs, such as 0.80, 0.85, 0.90, 0.95, etc.), the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked as a first exemplary image subsequence.

[0086] Specifically, in one embodiment, each exemplary image subsequence in the multiple exemplary image subsequences is marked as a to-be-processed image subsequence in turn, the to-be-processed image subsequence has image anomaly annotation data, and the image anomaly assessment data corresponding to the to-be-processed image subsequence analyzed by the intermediate image analysis neural network includes image anomaly characterization data corresponding to the to-be-processed image subsequence, and the image anomaly characterization data is used to reflect whether there is an abnormal image or the degree of image anomaly. Based on this, the step of determining a first exemplary image subsequence among the multiple exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the multiple exemplary image subsequences may include the following specific contents:

[0087] In the case where the image anomaly characterization data corresponding to the image subsequence to be processed is consistent with the image anomaly annotation data corresponding to the image subsequence to be processed, the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked to form a first exemplary image subsequence.

[0088] Specifically, in one embodiment, each image sequence or image subsequence in the exemplary building image sequence and the first exemplary image subsequence is sequentially marked as data for network optimization. Based on this, the step of performing network optimization processing on the intermediate image analysis neural network according to the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network may further include the following specific contents:

[0089] Extracting the shallow key information of the exemplary building contained in the network optimization data;

[0090] Mining the deep key information of the exemplary building corresponding to the network optimization data through the intermediate image analysis neural network (such as mapping the network optimization data to a feature space to form the corresponding deep key information of the exemplary building, or performing further knowledge extraction);

[0091] By means of the intermediate image analysis neural network, based on the shallow key information of the exemplary building and the deep key information of the exemplary building (such as after splicing, analysis and prediction are performed), the image anomaly assessment data corresponding to the network optimization data is analyzed;

[0092] Analyze the image anomaly analysis cost corresponding to the intermediate image analysis neural network based on the image anomaly assessment data corresponding to the network optimization data and the image anomaly annotation data of the network optimization data (i.e., the difference between the data);

[0093] According to the image anomaly analysis cost value, the intermediate image analysis neural network is subjected to network optimization processing to form a corresponding target image analysis neural network (i.e., the image anomaly analysis cost value is reduced to form a target image analysis neural network).

[0094] Specifically, in one implementation, the exemplary shallow key information of the building may further include the following specific contents:

[0095] Key information of building components corresponding to the building components possessed by the exemplary building images in the network optimization data, the key information of building components being formed by mining the type ratio information of the building components possessed by each exemplary building image in the network optimization data (such as the number ratio of building component type A being 1, the number ratio of building component type B being 2, the number ratio of building component type C being 3, the number ratio of building component type D being 4, and the number ratio of building component type E being 5), and the expression form of the key information of building components comprising a vector; image cumulative key information of a subsequence of building images in the network optimization data, the image cumulative key information being formed by mining the number of image frames and the amount of image data of the exemplary building images included in the building image subsequence, and the expression form of the image cumulative key information comprising a vector; and / or, component cumulative key information of the main building components among the building components possessed by the exemplary building images in the network optimization data, the component cumulative key information being formed by mining the number of components and the number of component types of the building components possessed by each exemplary building image in the network optimization data, and the expression form of the component cumulative key information comprising a vector (in addition, the expression form of the exemplary building deep key information may also comprise a vector).

[0096] Combination Figure 3 The embodiment of the present invention further provides a system for collecting building information data, which can be applied to the above-mentioned building information data collection platform. The system for collecting building information data can include the following software function modules:

[0097] An original image processing module is used to extract an original image set corresponding to the building to be analyzed, and split the original image set to form a plurality of original image subsets corresponding to the original image set;

[0098] A key information extraction module is used to mark each of the original image subsets as an image sequence to be analyzed, and extract shallow key information of the building corresponding to the image sequence to be analyzed, wherein the image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed;

[0099] A key information mining module is used to mine the deep key information of the building corresponding to the image sequence to be analyzed through a target image analysis neural network;

[0100] An abnormality analysis module is used to analyze the image abnormality analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network, and the image abnormality analysis data is used to reflect whether there is an abnormal image or the degree of image abnormality in the image of the building to be analyzed included in the image sequence to be analyzed;

[0101] The image acquisition module is used to screen the multiple original image subsets based on the image anomaly analysis data to acquire the target building image from the original image set.

[0102] Specifically, in one embodiment, the building information data collection system may further include the following software function modules:

[0103] An exemplary image extraction module is used to extract an exemplary building image sequence, wherein the number of image frames of the exemplary building image sequence does not exceed the number of pre-configured reference image frames, the exemplary building image sequence has image anomaly annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of the exemplary building image;

[0104] a sliding window segmentation processing module, configured to perform sliding window segmentation processing on the first number of building image subsequences according to a target sequence length, so as to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences;

[0105] An initial optimization processing module, used for performing initial optimization processing on the initial image analysis neural network according to the exemplary building image sequence to form an intermediate image analysis neural network;

[0106] a subsequence determination module, configured to perform image anomaly analysis on the plurality of exemplary image subsequences through the intermediate image analysis neural network, and determine a first exemplary image subsequence from the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences;

[0107] A network optimization processing module is used to perform network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network.

[0108] Specifically, in one embodiment, the building information data collection system may further include the following software function modules:

[0109] An inter-frame comparison and analysis module is used to perform inter-frame comparison and analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, an image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity;

[0110] The segmentation processing module is used to perform segmentation processing on the exemplary building image sequence according to the inter-frame difference identification information to form the first number of building image sub-sequences.

[0111] In summary, the present invention provides a method and system for collecting building information data, which can split the original image set to form multiple original image subsets; mark each original image subset as an image sequence to be analyzed, and extract the shallow key information of the building corresponding to the image sequence to be analyzed; mine the deep key information of the building corresponding to the image sequence to be analyzed through the target image analysis neural network; analyze the image abnormality analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network; based on the image abnormality analysis data, screen multiple original image subsets to collect the target building image in the original image set. Based on the aforementioned steps, since it no longer relies on the low-precision manual abnormality analysis, but adopts the neural network for analysis, and the basis of the neural network analysis includes the shallow key information of the building and the deep key information of the building, the analysis basis is more sufficient, therefore, the reliability of building data collection can be improved to a certain extent, thereby improving the problem of low reliability of building data collection existing in the prior art.

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for collecting building information data, characterized in that: include: Extracting an original image set corresponding to the building to be analyzed, and splitting the original image set to form a plurality of original image subsets corresponding to the original image set; Marking each of the original image subsets as an image sequence to be analyzed, and extracting shallow key information of the building corresponding to the image sequence to be analyzed, wherein the image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed; Mining out the deep key information of the building corresponding to the image sequence to be analyzed through the target image analysis neural network; By means of the target image analysis neural network, according to the shallow key information of the building and the deep key information of the building, image anomaly analysis data corresponding to the image sequence to be analyzed is analyzed, and the image anomaly analysis data is used to reflect whether there are abnormal images or the degree of image anomaly of the image of the building to be analyzed included in the image sequence to be analyzed; Based on the image anomaly analysis data, the plurality of original image subsets are screened to capture a target building image in the original image set; The method for collecting building information data also includes: An exemplary building image sequence is extracted, wherein the number of image frames of the exemplary building image sequence does not exceed the number of preconfigured reference image frames, the exemplary building image sequence has image anomaly annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of the exemplary building image; According to the target sequence length, performing sliding window segmentation processing on the first number of building image subsequences to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences; According to the exemplary building image sequence, the initial image analysis neural network is initially optimized to form an intermediate image analysis neural network; performing image anomaly analysis on the plurality of exemplary image subsequences by means of the intermediate image analysis neural network, and determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences; According to the exemplary building image sequence and the first exemplary image subsequence, performing network optimization processing on the intermediate image analysis neural network to form a target image analysis neural network corresponding to the intermediate image analysis neural network; Each image sequence or image subsequence in the exemplary building image sequence and the first exemplary image subsequence is sequentially marked as data for network optimization; The step of performing network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network includes: Extracting the shallow key information of the exemplary building contained in the network optimization data; By means of the intermediate image analysis neural network, deep key information of the exemplary building corresponding to the network optimization data is mined; By means of the intermediate image analysis neural network, based on the shallow key information of the exemplary building and the deep key information of the exemplary building, image anomaly assessment data corresponding to the network optimization data is analyzed; Analyze the image anomaly analysis cost value corresponding to the intermediate image analysis neural network based on the image anomaly assessment data corresponding to the network optimization data and the image anomaly labeling data of the network optimization data; According to the image abnormality analysis cost value, performing network optimization processing on the intermediate image analysis neural network to form a corresponding target image analysis neural network; The exemplary shallow key information of the building includes: Key information of building components corresponding to the building components of the exemplary building images in the network optimization data, the key information of building components being formed based on mining the type ratio information of the building components of each exemplary building image in the network optimization data, and the expression form of the key information of building components including vectors; image cumulative key information of a subsequence of building images in the network optimization data, the image cumulative key information being formed based on mining the number of image frames and the amount of image data of exemplary building images included in the subsequence of building images, and the expression form of the image cumulative key information comprising a vector; and / or The component cumulative key information of the main building components in the building components of the exemplary building images in the network optimization data is formed based on mining the component quantity and component type quantity of the building components in each exemplary building image in the network optimization data, and the expression form of the component cumulative key information includes a vector.

2. The method for collecting building information data according to claim 1, characterized in that: The method for collecting building information data also includes: Performing inter-frame comparison analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, an image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity; The exemplary building image sequence is segmented according to the inter-frame difference identification information to form the first number of building image sub-sequences.

3. The method for collecting building information data according to claim 1, characterized in that: Each of the plurality of exemplary image subsequences is marked as a to-be-processed image subsequence in turn, the to-be-processed image subsequence having image anomaly annotation data, the image anomaly assessment data of the to-be-processed image subsequence analyzed by the intermediate image analysis neural network comprising image anomaly characterization data corresponding to the to-be-processed image subsequence and an assessment possibility parameter for the image anomaly characterization data, the image anomaly characterization data being used to reflect whether there is an abnormal image or the degree of image anomaly; The step of determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences comprises: When the image anomaly characterization data corresponding to the image subsequence to be processed and the image anomaly annotation data corresponding to the image subsequence to be processed are consistent, and when the evaluation possibility parameter for the image subsequence to be processed is not less than a pre-configured reference evaluation possibility parameter, the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked as a first exemplary image subsequence.

4. The method for collecting building information data according to claim 1, characterized in that: Each exemplary image subsequence in the multiple exemplary image subsequences is marked as a to-be-processed image subsequence in turn, the to-be-processed image subsequence has image anomaly annotation data, and the image anomaly assessment data corresponding to the to-be-processed image subsequence analyzed by the intermediate image analysis neural network includes image anomaly characterization data corresponding to the to-be-processed image subsequence, and the image anomaly characterization data is used to reflect whether there is an abnormal image or the degree of image anomaly; The step of determining a first exemplary image subsequence from among the plurality of exemplary image subsequences based on the analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences comprises: In the case where the image anomaly characterization data corresponding to the image subsequence to be processed is consistent with the image anomaly annotation data corresponding to the image subsequence to be processed, the intermediate image analysis neural network is determined to have analysis reliability for the image subsequence to be processed, and the image subsequence to be processed is marked to form a first exemplary image subsequence.

5. A building information data collection system, characterized in that: include: An original image processing module is used to extract an original image set corresponding to the building to be analyzed, and split the original image set to form a plurality of original image subsets corresponding to the original image set; A key information extraction module is used to mark each of the original image subsets as an image sequence to be analyzed, and extract shallow key information of the building corresponding to the image sequence to be analyzed, wherein the image sequence to be analyzed includes multiple frames of building images to be analyzed corresponding to the building to be analyzed; A key information mining module is used to mine the deep key information of the building corresponding to the image sequence to be analyzed through a target image analysis neural network; An abnormality analysis module is used to analyze the image abnormality analysis data corresponding to the image sequence to be analyzed based on the shallow key information of the building and the deep key information of the building through the target image analysis neural network, and the image abnormality analysis data is used to reflect whether there is an abnormal image or the degree of image abnormality in the image of the building to be analyzed included in the image sequence to be analyzed; An image acquisition module, configured to screen the plurality of original image subsets based on the image anomaly analysis data, so as to acquire a target building image from the original image set; The building information data collection system also includes: An exemplary image extraction module is used to extract an exemplary building image sequence, wherein the number of image frames of the exemplary building image sequence does not exceed the number of pre-configured reference image frames, the exemplary building image sequence has image anomaly annotation data, the exemplary building image sequence includes a first number of building image subsequences, and each of the first number of building image subsequences includes an adjacent frame of the exemplary building image; a sliding window segmentation processing module, configured to perform sliding window segmentation processing on the first number of building image subsequences according to a target sequence length, so as to output a plurality of exemplary image subsequences corresponding to the exemplary building image sequence, each of the exemplary image subsequences including at least one adjacent building image subsequence in the first number of building image subsequences; An initial optimization processing module, used for performing initial optimization processing on the initial image analysis neural network according to the exemplary building image sequence to form an intermediate image analysis neural network; a subsequence determination module, configured to perform image anomaly analysis on the plurality of exemplary image subsequences through the intermediate image analysis neural network, and determine a first exemplary image subsequence from the plurality of exemplary image subsequences based on analysis reliability of the intermediate image analysis neural network for the plurality of exemplary image subsequences; A network optimization processing module is used to perform network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network.

6. The building information data collection system according to claim 5, characterized in that: The building information data collection system also includes: An inter-frame comparison and analysis module is used to perform inter-frame comparison and analysis on the exemplary building image sequence to determine inter-frame difference identification information in the exemplary building image sequence, wherein in a sequence position corresponding to the inter-frame difference identification information, an image similarity between two frames of exemplary building images corresponding to the sequence position is less than or equal to a pre-configured reference image similarity; a segmentation processing module, configured to perform segmentation processing on the exemplary building image sequence according to the inter-frame difference identification information to form the first number of building image subsequences; Each image sequence or image subsequence in the exemplary building image sequence and the first exemplary image subsequence is sequentially marked as data for network optimization; The step of performing network optimization processing on the intermediate image analysis neural network based on the exemplary building image sequence and the first exemplary image subsequence to form a target image analysis neural network corresponding to the intermediate image analysis neural network includes: Extracting the shallow key information of the exemplary building contained in the network optimization data; By means of the intermediate image analysis neural network, deep key information of the exemplary building corresponding to the network optimization data is mined; By means of the intermediate image analysis neural network, based on the shallow key information of the exemplary building and the deep key information of the exemplary building, image anomaly assessment data corresponding to the network optimization data is analyzed; Analyze the image anomaly analysis cost value corresponding to the intermediate image analysis neural network based on the image anomaly assessment data corresponding to the network optimization data and the image anomaly labeling data of the network optimization data; According to the image abnormality analysis cost value, performing network optimization processing on the intermediate image analysis neural network to form a corresponding target image analysis neural network; The exemplary shallow key information of the building includes: Key information of building components corresponding to the building components of the exemplary building images in the network optimization data, the key information of building components being formed based on mining the type ratio information of the building components of each exemplary building image in the network optimization data, and the expression form of the key information of building components including vectors; image cumulative key information of a subsequence of building images in the network optimization data, the image cumulative key information being formed based on mining the number of image frames and the amount of image data of exemplary building images included in the subsequence of building images, and the expression form of the image cumulative key information comprising a vector; and / or The component cumulative key information of the main building components in the building components of the exemplary building images in the network optimization data is formed based on mining the component quantity and component type quantity of the building components in each exemplary building image in the network optimization data, and the expression form of the component cumulative key information includes a vector.

Images