A method, system and device for extracting drawing space information

Abstract

Embodiments of the present specification provide a method, system, and device for extracting spatial information from a drawing, wherein the method includes: obtaining a frame of a target drawing; performing label recognition based on the frame to determine initial spatial information; verifying the initial spatial information based on a preset spatial information database to obtain a verification result; splitting the initial spatial information based on the verification result to obtain spatial description information; and obtaining target spatial information of the target drawing based on the spatial description information.

Description

Technical Field

[0001] This specification relates to the field of drawing recognition, and in particular to a drawing space information extraction method, system and device. Background Art

[0002] In the field of engineering construction, spatial information is often included in the sub-item names, drawing titles, and descriptions of various materials on drawings, requiring users to read them. To optimize the management of digital platforms in the engineering construction field, certain technologies are needed to extract this information. Typically, users are required to manually enter spatial information. For example, fields related to spatial information are provided and manually filled in by the user. However, this manual entry method still relies on users to read relevant materials, lacks intelligence, and is difficult to process large amounts of data, failing to meet the needs of drawing management.

[0003] Therefore, it is necessary to provide a drawing space information extraction method, system and device to achieve efficient drawing space information management and retrieval. Summary of the Invention

[0004] One or more embodiments of the present specification provide a method for extracting spatial information of a drawing, the method comprising: obtaining a frame of a target drawing; the frame comprising label information; identifying the label based on the frame to determine initial spatial information; comparing the initial spatial information based on a preset spatial information database to obtain a comparison result; splitting the initial spatial information based on the comparison result to obtain spatial description information; and obtaining target spatial information of the target drawing based on the spatial description information.

[0005] One or more embodiments of the present specification also provide a drawing space information extraction system, the system comprising: a first acquisition module, configured to acquire a frame of a target drawing; the frame comprises label information; an identification module, configured to perform label identification based on the frame to determine initial spatial information; a comparison module, configured to compare the initial spatial information based on a preset spatial information database to obtain a comparison result; a second acquisition module, configured to split the initial spatial information according to the comparison result to obtain spatial description information; and a third acquisition module, configured to acquire target spatial information of the target drawing based on the spatial description information.

[0006] One or more embodiments of this specification also provide a drawing space information extraction device, which includes at least one storage medium and at least one processor; the at least one storage medium is used to store computer instructions; the at least one processor is used to execute the computer instructions to implement the drawing space information extraction method described in the above embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein:

[0008] Figure 1 is a schematic diagram of an application scenario of a drawing space information extraction system according to some embodiments of this specification;

[0009] Figure 2 is an exemplary flow chart of a drawing space information extraction method according to some embodiments of this specification;

[0010] Figure 3 is an exemplary flow chart for determining target space information according to other embodiments of this specification;

[0011] Figure 4 is an exemplary flow chart of verifying target spatial information based on a target drawing according to some embodiments of this specification;

[0012] Figure 5 is a partial schematic diagram of a frame area of ​​a target drawing according to some embodiments of this specification;

[0013] Figure 6 It is an exemplary module diagram of a drawing space information extraction system according to some embodiments of this specification. DETAILED DESCRIPTION

[0014] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0015] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0016] As used in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but also include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0017] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0018] Figure 1 It is a schematic diagram of an application scenario of a drawing space information extraction system according to some embodiments of this specification.

[0019] In some embodiments, as Figure 1 As shown, the application scenario 100 of the drawing space information extraction system includes a target drawing 110, a network 120, a terminal 130, a processor 140, and a storage device 150.

[0020] The target drawing 110 is a schematic diagram for which spatial information extraction is required. For example, the target drawing 110 may be a schematic diagram of an architecture, a renovation, a machinery manufacturing, or a clothing design. The types of target drawings 110 listed above are merely examples and are not intended to be limiting. The target drawing 110 may be any drawing for which spatial information extraction is required.

[0021] In some embodiments, the target drawing 110 may be a physical drawing (such as a paper drawing). In other embodiments, the target drawing 110 may also be an electronic drawing. For more information about the target drawing, see Figure 2 and its related parts.

[0022] The network 120 can connect various components in the application scenario 100 of the drawing space information extraction system and / or connect other components outside the application scenario 100. In some embodiments, one or more components of the application scenario 100 of the drawing space information extraction system (e.g., the terminal 130, the processor 140, and the storage device 150) can be connected and / or communicate with each other via the network 120. For example, the terminal 130 can send the label of the target drawing 110 to the processor 140 via the network 120.

[0023] The terminal 130 can provide functional components related to user interaction and can implement user interaction functions (e.g., providing or displaying information and data to the user). The user can refer to the administrator of the target drawing 110 or the person who searches for or uses the target drawing 110. By way of example only, the terminal 130 can be a mobile device, a tablet computer, a laptop computer, a desktop computer, or any other device with input and / or output functions, or any combination thereof.

[0024] The processor 140 is capable of processing information and / or data related to the drawing space information extraction system to perform one or more functions described in this specification. In some embodiments, the processor 140 can obtain a frame of a target drawing; the frame includes label information; the label is identified based on the frame to determine the initial spatial information; the initial spatial information is compared based on a preset spatial information database to obtain a comparison result; based on the comparison result, the initial spatial information is split to obtain spatial description information; based on the spatial description information, the target spatial information of the target drawing is obtained. Detailed descriptions of related content can be found later (e.g. Figure 2 、 Figure 3 etc.) related descriptions.

[0025] In some embodiments, the processor 140 may include a central processing unit (CPU), a digital signal processor (DSP), a microcontroller unit (MCU), a computer, a user console, or the like, or any combination thereof. In some embodiments, the processor 140 may include a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 140 may be local or remote. In some embodiments, the processor 140 may be implemented on a cloud platform. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, or the like, or any combination thereof.

[0026] The storage device 150 is capable of storing data, instructions, and / or any other information. In some embodiments, the storage device 150 may store data obtained from the terminal 130, the processor 140, etc., such as the target drawing 110, drawing space information, etc. In some embodiments, the storage device 150 may include a mass storage device, a removable memory, a volatile read-write memory, a read-only memory (ROM), etc., or any combination thereof. In some embodiments, the storage device 150 may be executed on a cloud platform. In some embodiments, the storage device 150 may be connected to the network 120 to communicate with one or more other components of the application scenario 100 of the drawing space information extraction system (e.g., the terminal 130, the processor 140, etc.). In some embodiments, the storage device 150 may be part of the processor 140.

[0027] It is worth noting that the application scenario 100 of the drawing space information extraction system is provided for illustrative purposes only and is not intended to limit the scope of this specification. For those skilled in the art, various changes and modifications can be made based on the description of this specification. For example, the application scenario 100 of the drawing space information extraction system can also include a database, an information source, etc. For another example, the application scenario 100 of the drawing space information extraction system can be implemented on other devices to achieve similar or different functions. However, these changes and modifications will not deviate from the scope of this specification.

[0028] Figure 2 is an exemplary flow chart of a drawing space information extraction method according to some embodiments of this specification. In some embodiments, process 200 may be executed by a processing device (eg, processor 140). Figure 2 As shown, the process 200 includes the following steps.

[0029] Step 210: Acquire a frame of the target drawing. In some embodiments, the frame includes label information.

[0030] The drawing frame is a boundary line that identifies the boundary of the drawing. In some embodiments, the target drawing includes a drawing area and a label area. The drawing area is used to draw a specific design drawing, and the label area is used to identify the label information.

[0031] Drawing label information is a label used to identify drawing information in a drawing.

[0032] The drawing label information may include various information related to the drawing. For example only, the drawing label information may include the drawing title, project name, drawing number, version, drawing unit, etc.

[0033] Generally speaking, the specific drawing content (i.e. the drawing area) is placed in the middle of the drawing frame, and the drawing label information is placed in the inner corners of the drawing frame.

[0034] In some embodiments, the processor 140 may identify a frame in a target drawing. For example, the processor 140 may annotate the frame on the target drawing using an annotation tool to obtain the frame of the target drawing. Exemplary annotation tools may include various CAD annotation software.

[0035] In some embodiments, when the target drawing is a physical drawing, the physical drawing can be converted into an electronic drawing by photographing, scanning, etc., and then the frame of the target drawing can be obtained.

[0036] Step 220 : performing image label recognition based on the image frame to determine initial spatial information.

[0037] Initial spatial information refers to unverified spatial information extracted from the target drawing. In some embodiments, processor 140 may further process the initial spatial information to obtain more accurate and complete target spatial information. Spatial information refers to information / data describing the spatial attribution of a building as depicted in the drawing. For example, spatial information may indicate the specific floor and / or building of the building depicted in the drawing.

[0038] Figure 5 It is a partial schematic diagram of a frame area of ​​a target drawing according to some embodiments of this specification.

[0039] In some embodiments, spatial information may include information related to building keywords, floor keywords, building numbers, and floor numbers. The building keywords and building numbers may be referred to as building information related to the building in the drawing; the floor keywords and floor numbers may be referred to as floor information related to the floor in the drawing. For example, the title of the target drawing may include information related to the building keyword, floor keyword, building number, and floor number.

[0040] Building keywords refer to the identification information related to the indicated building in the target drawing. For example, see Figure 5 , the building keyword 540 may include "building" or the like. In some embodiments, the building keyword may also include other identification information that may indicate a building, such as "building," "building number" (e.g., "Building 1," "Building 2"), etc. In some embodiments, the processor 140 may determine whether the image tag information is related to the building based on the building keyword. For example, when a certain image tag information includes building keywords such as "building," "building," and "building number," the processor 140 may determine that the image tag information is related to the building.

[0041] Floor keywords refer to the identification information related to the indicated floor in the target drawing. For example, see Figure 5 , the floor keyword 550 may include "floor" or the like. In some embodiments, the floor keyword may also include other identification information that may indicate a floor, such as "floor," "F," or the like. In some embodiments, the processor 140 may determine whether the image tag information is related to a floor based on the floor keyword. For example, if a certain image tag information includes floor keywords such as "floor," "floor," or "F," the processor 140 may determine that the image tag information is related to a floor.

[0042] The building number is data / information used to distinguish different buildings. In some embodiments, the building number includes numbers, hyphens, custom buildings, etc.

[0043] The floor number is data / information for distinguishing different floors. In some embodiments, the floor number includes numbers, connecting symbols, user-defined floors, prefixes, supplementary words, etc.

[0044] Numbers are used to indicate specific buildings and / or floors, including Arabic numerals (such as 1, 2, 3, etc.), Chinese numerals (such as one, two, three, etc.), Roman numerals, etc. Figure 5 As shown, the numbers include a first number 510 indicating a specific floor and a second number 520 indicating a specific building, etc. For more description of the first number and the second number, please refer to step 250.

[0045] Connecting symbols are used to connect multiple different information. Figure 5 As shown, the connecting symbol 530 can connect different building numbers and / or floor numbers, different numbers, etc. Connecting symbols include separators (such as ",",","and ...,and,"and,"and,"and,"and,"and,and,"and,"and,"and,and,"and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,"and,"and,and,

[0046] Custom buildings are words that indicate non-standard descriptions of buildings. For example, custom buildings may include "basement, garage, gatehouse," etc.

[0047] Custom floor is a term that indicates a non-standardized description of a floor. Figure 5 As shown, the custom floor 560 can be a "device floor". In some embodiments, the custom floor can also include types such as "roof, machine room floor, roof, top floor" and the like.

[0048] Prefixes and supplementary words are used to explain and supplement special floors. For example, prefixes such as "-", "B" can indicate a basement floor; supplementary words such as "mezzanine", "half floor", "M" can indicate a mezzanine floor, etc.

[0049] In some embodiments, the processor 140 may determine the initial spatial information in a variety of ways. For example, the processor 140 may identify the image frame using image recognition (e.g., optical character recognition, OCR) and other technologies, obtain the image label information such as the image name of the image frame, and extract the building keyword, floor keyword, building number, floor number, etc. from the image label information as the initial spatial information. For another example, the processor 140 may identify the image label of the image frame using machine vision technology, image recognition technology, and other methods to extract the initial spatial information.

[0050] Step 230 : Compare the initial spatial information based on a preset spatial information database to obtain a comparison result.

[0051] Spatial information database refers to a database system used to store, manage and retrieve spatial information of drawings.

[0052] In some embodiments, the spatial information in the spatial information database can be obtained by collecting spatial information of a large amount of existing drawings and performing standardization processing.

[0053] Standardization processing refers to processing the spatial information related to buildings, floors, etc. in the drawings based on preset rules. An exemplary preset rule may include: arranging different buildings and their corresponding floors alternately. For example, the spatial information related to buildings, floors, etc. in the extracted drawings is arranged in the manner shown in formula (1):

[0054] (α×n1+β×n2)×n3(1)

[0055] Here, α represents the standardized representation of building information, β represents the standardized representation of floor information, and n1, n2, and n3 represent the number of buildings, the number of floors, and the number of occurrences of building and / or floor information, respectively. For example, "Buildings 1-3, floors 1-3, equipment floor, and Buildings 5 ​​and 7, floors 2-5 floor plan" can represent standardized spatial information.

[0056] In some embodiments, the standardized representation of building information may include: (number | building keyword + hyphen) + (number | building keyword + hyphen) + ... + (number | building keyword + hyphen). For example, the standardized representation of building information may be: Buildings 1-3, 5, 7, 9-10. Building keywords include custom building names such as "basement," "garage," and "gateway."

[0057] In some embodiments, the standardized representation of floor information may include: (number | floor keyword + hyphen) + (number | floor keyword + hyphen) + ... + (number | floor keyword + hyphen). For example, the standardized representation of floor information may include: 1st to 3rd, 5th, 7th, 9th, and equipment floor. Floor keywords include custom floors such as "rooftop, equipment floor, computer room floor, roof, and top floor."

[0058] The above description of the standardization process is only an example, and those skilled in the art may also set other standardization processes according to different drawing spatial information extraction requirements. For example, the spatial information database may include a variety of spatial information obtained through different standardization processes.

[0059] In some embodiments, the processor 140 may match the initial spatial information of the target drawing with the spatial information in the spatial information database, and use the matching result as the comparison result. The matching method includes text matching, etc. For example, the processor 140 may match the building keyword, floor keyword, building number, and floor number in the key information of the frame of the target drawing with the spatial information in the spatial information database for comparison.

[0060] In some embodiments, the initial spatial information is compared against a preset spatial information database, and the comparison results obtained include various types. For example, when the floor keywords (including custom floors) and building keywords (including custom buildings) in the initial spatial information successfully match the floor keywords (including custom floors) and building keywords (including custom buildings) in the spatial database, the processor 140 may determine that the initial spatial information contains spatial information related to both floors and buildings, and this result is referred to as a first comparison result. When the floor keywords (including custom floors) in the initial spatial information successfully match the floor keywords (including custom floors) in the spatial database, the processor 140 may determine that the initial spatial information contains spatial information related to floors, and this result is referred to as a second comparison result. When the floor keywords (including custom floors) in the initial spatial information fail to match the floor keywords (including custom floors) in the spatial database, the processor 140 may determine that the initial spatial information does not contain spatial information related to floors, and further determine that the target drawing does not contain spatial information, and this result is referred to as a third comparison result.

[0061] In some embodiments, when the comparison result satisfies the first comparison result, the processor 140 may determine the target space information of the target drawing by the method described in steps 240 to 250 .

[0062] In some embodiments, when the comparison result satisfies the second comparison result, the processor 140 may determine that the target drawing's name only includes floor information, but not building information. The target floor information in the target space information may be obtained from the drawing name, and the target building information may be obtained from the sub-item information. The processor 140 may determine the target space information of the target drawing using the method described in process 300. For more information on related content, please refer to Figure 3 .

[0063] In some embodiments, when the comparison result satisfies the third comparison result, the processor 140 may determine that the target drawing does not include spatial information and terminate the process.

[0064] Step 240: split the initial spatial information according to the comparison result to obtain spatial description information.

[0065] The space description information is data / information describing buildings, floors, etc. in the target drawing. In some embodiments, the space description information includes building space description information describing buildings and floor space description information describing floors.

[0066] In some embodiments, when the comparison result meets the first preset condition, the processor 140 can split the initial spatial information to obtain multiple spatial information combinations; for each spatial information combination, the processor 140 can obtain floor spatial description information and building spatial description information based on a preset method.

[0067] The first preset condition may include that the comparison result is a first comparison result. For more information about the first comparison result, see step 230.

[0068] A spatial information combination refers to the combination of building spatial description information and floor spatial description information obtained after decomposing the initial spatial information. For example, processor 140 may decompose the initial spatial information "Buildings 1-3, floors 1-3, equipment floor, and Buildings 5-7, floors 2-5 floor plan" into spatial information combination 1: "Buildings 1-3, floors 1-3, equipment floor" and spatial information combination 2: "Buildings 5-7, floors 2-5."

[0069] In some embodiments, if a building keyword or a custom building is preceded by a floor keyword or a custom floor, the processor 140 may use the connection symbol connecting the building keyword or custom building and the preceding floor keyword or custom floor as a dividing point to split the initial spatial information. Continuing to use the aforementioned initial spatial information "Buildings 1-3, 1-3 floors, equipment floor and Buildings 5-7, 2-5 floors architectural plan", where the building keyword of "Building 5" includes the floor keyword "floor" of "1-3 floors" and the custom floor "equipment floor", the processor 140 may use the connection symbol "and" as a dividing point to split the initial spatial information into spatial information combination 1: "Buildings 1-3, 1-3 floors, equipment floor" and spatial information combination 2: "Buildings 5-7, 2-5 floors".

[0070] In some embodiments, the preset method includes: extracting information related to floors based on the spatial information combination to obtain floor space description information; extracting information related to buildings based on the spatial information combination to obtain building space description information.

[0071] In some embodiments, for each of the multiple spatial information combinations, extracting floor-related information includes deleting the building keyword and the characters preceding it, and extracting the floor keyword and floor number from the remaining characters. For example, for spatial information combination 1: "Buildings 1-3, Floors 1-3, Equipment Floor," processor 140 may delete the building keyword "building" and the characters preceding it, "1-3," and extract the floor-related information "Floors 1-3, Equipment Floor" as the floor spatial description information.

[0072] In some embodiments, for each of the multiple spatial information combinations, extracting the building-related information includes: deleting the rightmost characters after the building keyword and the custom building, and extracting the building keyword and building number from the remaining characters. For example, for spatial information combination 1: "Buildings 1-3, Floors 1-3, Equipment Floor," processor 140 may delete the rightmost characters "1-3 Floors, Equipment Floor" after the building keyword "Building" and extract the building-related information "Buildings 1-3" as the building spatial description information.

[0073] Step 250: Acquire target space information of the target drawing based on the space description information.

[0074] The target spatial information refers to the spatial information corresponding to buildings and floors extracted from the target drawings. For example, for the initial spatial information "Buildings 1-3, floors 1-3, equipment floor, and Buildings 5-7, floors 2-5," the corresponding target spatial information includes: "Building 1: 1st floor, 2nd floor, 3rd floor, equipment floor," "Building 2: 1st floor, 2nd floor, 3rd floor, equipment floor," "Building 3: 1st floor, 2nd floor, 3rd floor, equipment floor," "Building 5: 2nd floor, 3rd floor, 4th floor, 5th floor," and "Building 7: 2nd floor, 3rd floor, 4th floor, 5th floor."

[0075] In some embodiments, for each of multiple spatial information combinations, the processor 140 can perform keyword extraction based on the floor space description information and the building space description information to obtain the target floor information and the target building information; combine the target floor information and the target building information to obtain the target space information of the target drawing.

[0076] In some embodiments, the processor 140 can extract the first number and the custom floor in the floor space description information; determine the first floor information based on the adjacent connecting words of the first number; determine the second floor information based on the consecutive words between the first numbers; determine the target floor information based on the first floor information and the second floor information, and use the custom floor as supplementary floor information.

[0077] The first digit refers to a digit describing a specific floor number, wherein the first digit may include digits in various forms, such as Arabic numerals (such as 1, 2, 3, etc.), Chinese numerals (such as one, two, three, etc.), Roman numerals, etc.

[0078] Adjacent connectives refer to words that are immediately adjacent to a number (including the first and second numbers). In some embodiments, the adjacent connectives of the first number include prefixes, supplementary words, floor keywords, etc. Taking the floor space description information "2nd floor below ground" and "1st floor below ground" as an example, adjacent connectives include "underground," "floor," "underground," and "floor." For more information about prefixes, supplementary words, and floor keywords, see step 220.

[0079] The first floor information refers to floor information determined based on a first number and adjacent conjunctions to the first number.

[0080] In some embodiments, the processor 140 may process adjacent conjunctions of the first number and replace the first number to obtain first floor information. An exemplary processing method includes converting the adjacent conjunctions into standardized prefixes and / or supplementary words and / or floor keywords, and combining the standardized prefixes and / or supplementary words and / or floor keywords with the first number to obtain the first floor information. For example, taking "-1 floor," "-2F," "1st floor underground," and "1st floor basement" as examples, the processor 140 may convert adjacent conjunctions such as "-," "-," "underground," and "basement" representing underground into a standardized prefix "-" (or other prefixes as needed), convert floor keywords such as "floor," "floor," and "F" into a standardized floor keyword "floor," and then combine the standardized prefix "-" and the standardized floor keyword "floor" with the corresponding first number to obtain the first floor information: "-1 floor," "-2 floor," "-1 floor," and "-1 floor." The standardized prefixes and / or supplementary words and / or floor keywords may be preset in advance.

[0081] In some embodiments, when the first number in the floor space description information does not have an adjacent connective, the processor 140 may determine that the first floor information of the floor space description information is empty.

[0082] In some embodiments, the processor 140 may store the first floor information in a temporary floor database.

[0083] The second floor information refers to the floor information determined based on the first number and the consecutive words between the first number. For more information about the consecutive words, please refer to step 220.

[0084] In some embodiments, when there are no consecutive words between the first numbers, the processor 140 may directly determine the first numbers and the floor keyword as the second floor information.

[0085] In some embodiments, when there is a consecutive word between the first numbers, the processor 140 may supplement the remaining numbers between the floor numbers on both sides of the consecutive word, and use the supplemented result and the standardized floor keyword as the second floor information. For example, taking "floors 1 to 4" as an example, there is a consecutive word "to" indicating the consecutive situation of floors between the first numbers "1" and "4". The processor 140 may supplement the numbers "2" and "3" between the first numbers "1" and "4", and combine the supplemented result with the standardized floor keyword "floor" as the second floor information: "Floor 1", "Floor 2", "Floor 3", "Floor 4".

[0086] In some embodiments, the processor 140 may store the second floor information in the temporary floor database.

[0087] In some embodiments, the processor 140 may merge the first floor information and the second floor information in the temporary floor database as the target floor information.

[0088] The supplementary floor information refers to the information that supplements and explains the target floor information. In some embodiments, the processor 140 may directly use the custom floor information as the supplementary floor information and store it in the temporary floor database.

[0089] In some embodiments, the processor 140 may also merge the supplementary floor information, the first floor information, and the second floor information as the target floor information.

[0090] In some embodiments, for each of the multiple combinations of spatial information, the processor also may perform a duplicate removal process (i.e., remove duplicate floor information) on the target floor information in the spatial information combination in the temporary floor database to improve the accuracy of the subsequent obtained target spatial information.

[0091] In some embodiments, the processor 140 may extract the second number and the custom building in the building spatial description information; determine the first building information based on the adjacent connecting words of the second number; determine the second building information based on the consecutive words between the second numbers; determine the target building information based on the first building information and the second building information, and use the custom building as the supplementary floor information.

[0092] The second number refers to the number that describes the specific building number. Among them, the second number may include various forms of numbers, such as natural numbers (such as 1, 2, 3, etc.), Chinese numerals (such as one, two, three, etc.), Roman numerals, etc.

[0093] In some embodiments, the adjacent conjunctions of the second number include building keywords. Taking the building space description information "Building 3" and "Buildings 1 to 4" as examples, the adjacent conjunctions include "Building" and "Block", etc. For more descriptions of the building keywords, refer to step 220.

[0094] The first building information refers to the building information determined based on the second number and the adjacent conjunctions of the second number.

[0095] In some embodiments, the processor 140 can process the adjacent conjunctions of the second number and obtain the first building information after replacing the second number. Among them, exemplary processing methods include converting the adjacent conjunctions into standardized building keywords and combining the standardized building keywords with the second number to obtain the first building information. For example, taking "Building 3", "Buildings 1 to 4", "Building No. 7", etc. as examples, the processor 140 can convert the different building keywords such as "Building", "Block", and "Building No." into the standardized building keyword "Building" (it can also be converted into other forms of building keywords according to needs), and then combine the standardized building keyword "Building" with the corresponding second number to obtain the first building information: "Building 3", "Buildings 1 to 4", "Building 7", etc. Among them, the standardized building keyword can be preset in advance.

[0096] In some embodiments, the processor 140 can store the first building information in the temporary floor database.

[0097] The second building information refers to the building information determined based on the second number and the continuous words between the second numbers. For more descriptions of the continuous words, refer to step 220.

[0098] In some embodiments, when there are no continuous words between the second numbers, the processor 140 can directly determine the second number and the building keyword as the second building information.

[0099] In some embodiments, when there are continuous words between the second numbers, the processor 140 can fill in the remaining numbers between the building numbers on both sides of the continuous words, and use the filled result and the standardized building keyword as the second building information. For example, taking "from 1 to 4 buildings" as an example, there is a continuous word "to" indicating the continuous situation of buildings between the second numbers "1" and "4". The processor 140 can fill in the numbers "2" and "3" between the second numbers "1" and "4", and combine the filled result with the standardized building keyword "Building" as the second building information: "Building 1", "Building 2", "Building 3", "Building 4".

[0100] In some embodiments, the processor 140 can store the second building information in the temporary building database.

[0101] In some embodiments, the processor 140 may combine the first building information and the second building information in the temporary building database as the target building information.

[0102] The supplementary building information refers to information that supplements the target building information. In some embodiments, the processor 140 can directly use the customized building information as the supplementary building information and store it in the temporary building database.

[0103] In some embodiments, the processor 140 may further combine the supplementary building information with the first building information and the second building information as the target building information.

[0104] In some embodiments, for each of the multiple spatial information combinations, the processor 140 may further perform deduplication processing on the target building information in the spatial information combination in the temporary building database (i.e., remove duplicate building information) to improve the accuracy of the target spatial information subsequently acquired.

[0105] In some embodiments, for each of multiple spatial information combinations, the processor 140 may combine the target floor information and target building information in the spatial information combination to obtain the target spatial information of the spatial information combination. For example, if the target floor information of a spatial information combination includes "1st floor," "2nd floor," and "equipment floor," and the target building information includes "Building 1," "Building 2," and "Building 3," the processor 140 may combine each piece of target floor information with each piece of target building information, and then merge the combination results for the same building to obtain the target spatial information of the spatial information combination: "Building 1: 1st floor, 2nd floor, equipment floor," "Building 2: 1st floor, 2nd floor, equipment floor," and "Building 3: 1st floor, 2nd floor, equipment floor."

[0106] In some embodiments, for each of the multiple spatial information combinations of the target drawing, the processor 140 can separately execute the method described in step 250 (for example, simultaneously or separately) to obtain the target spatial information of each spatial information combination, and then merge the target spatial information of each spatial information combination as the target spatial information of the target drawing.

[0107] The drawing space information extraction method described in some embodiments of this specification can intelligently and automatically identify the spatial information in the drawing, with high recognition accuracy and good versatility, and can also accurately identify customized floor / building information; at the same time, when the drawing text is not expressed in a standardized manner (for example, there are multiple building / floor keywords, etc.), it can also be accurately identified and converted into a standardized expression to facilitate the query and management of the drawing space information; in addition, for the spatial information that is implicit and not directly expressed in the drawing text (for example, for continuous floor information 1 to 4 floors, where floors 2 and 3 are implicit spatial information), it can also be accurately supplemented and identified, further improving the drawing space information management effect.

[0108] Figure 3 This is an exemplary flowchart for determining target spatial information according to other embodiments of this specification. In some embodiments, when the comparison result satisfies a second preset condition, processor 140 may execute process 300 to determine the target spatial information of the target drawing. That is, after executing step 230, processor 140 may determine whether the comparison result satisfies the second preset condition. If the comparison result satisfies the second preset condition, step 300 is executed. The second preset condition includes the comparison result satisfying a second comparison result. For more information on the second comparison result, see step 230.

[0109] In some embodiments, process 300 may be performed by a processing device (e.g., processor 140). Figure 3 As shown, the process 300 includes the following steps.

[0110] Step 310: split the initial space information to obtain floor space description information.

[0111] In some embodiments, the processor 140 may split the initial space information to obtain the floor space description information using the same method as in step 240. Detailed description of the relevant content can be found in step 240.

[0112] Step 320: extract keywords based on the floor space description information to obtain target floor information.

[0113] In some embodiments, the processor 140 may extract keywords from the floor space description information to obtain the target floor information using the same method as in step 250. Detailed description of the relevant content can be found in step 250.

[0114] Step 330 : Obtain sub-item information of the frame, and based on the sub-item information, obtain sub-item spatial information.

[0115] Sub-item information refers to the sub-item name of the drawing. Sub-item refers to the next level of sub-item or sub-item in the drawing.

[0116] Sub-item space information refers to the building keyword, building number and other related information in the sub-item name.

[0117] In some embodiments, the processor 140 may determine the sub-item spatial information using the same method used to extract the initial spatial information. For example, the processor 140 may identify the frame using image recognition (e.g., OCR technology), obtain the sub-item name of the frame, and extract the building keyword, building number, etc. from the sub-item name as the sub-item spatial information using methods such as regular expression matching. For another example, the processor 140 may identify the image label of the frame using machine vision technology, image recognition technology, etc., and extract the sub-item spatial information.

[0118] In some embodiments, the processor 140 may further compare the sub-item spatial information based on a preset spatial information database.

[0119] In some embodiments, when the comparison result is that the sub-item spatial information successfully matches the building keyword (including custom buildings) in the spatial information database, it means that the sub-item spatial information includes building spatial information, and the processor 140 continues to execute steps 340 to 360 to determine the target spatial information of the target drawing.

[0120] In some embodiments, when the comparison result is that the sub-item spatial information fails to match the building keyword (including custom buildings) in the spatial information database, it means that the sub-item spatial information does not include building spatial information, and the processor 140 ends the process.

[0121] Step 340: Split the sub-item spatial information to obtain building spatial description information.

[0122] In some embodiments, processor 140 can separate the sub-item spatial information by custom building and / or building keyword and / or separator. For example, for the sub-item spatial information "Buildings 1-3, Buildings 7 and 8," the processor can separate the sub-item spatial information based on the separator "," and the building keyword "building," obtaining the building spatial information: "Buildings 1-3," "Building 7," and "Building 8." For more information about building spatial description information, see step 240.

[0123] Step 350: extract keywords based on the building space description information to obtain target building information.

[0124] In some embodiments, the processor 140 may extract keywords from the building space description information based on the same method as step 250 to obtain target building information. Detailed description of the relevant content can be found in step 250.

[0125] Step 360: Combine the target floor information and the target building information to obtain the target space information of the target drawing.

[0126] In some embodiments, the processor 140 may combine the target floor information and the target building information to obtain the target space information of the target drawing using the same method as step 250. Detailed description of the relevant content can be found in step 250.

[0127] In some embodiments, to further improve the accuracy of the extracted target spatial information, the processor 140 may also verify the extracted target spatial information. For example, the processor 140 may compare the extracted target spatial information with the spatial information in the spatial information database. If the comparison results are consistent, the extracted target spatial information is considered accurate and the target spatial information corresponding to the target drawing (i.e., the building and floor information associated with the drawing) is obtained. If the comparison results are inconsistent, the target spatial information is considered incorrect and needs to be modified or further verified (such as manually verified or modified).

[0128] In some embodiments, the processor 140 may also verify the target space information based on the target drawing.

[0129] Figure 4 This is an exemplary flow chart of verifying target space information based on target drawings according to some embodiments of this specification. In some embodiments, process 400 can be executed by a processing device (e.g., processor 140). Figure 4 As shown, process 400 includes the following steps.

[0130] Step 410: perform image segmentation based on the target drawing to obtain one or more building area drawings.

[0131] A building area drawing refers to a drawing obtained by dividing different buildings in a target drawing into one or more separate buildings.

[0132] In some embodiments, the processor 140 may perform image segmentation on the target drawing based on a variety of image segmentation methods, such as grayscale threshold segmentation, region segmentation, and edge segmentation.

[0133] Step 420: Determine auxiliary building information based on one or more building area drawings.

[0134] Auxiliary building information is building information extracted from building area drawings.

[0135] In some embodiments, the processor 140 may determine the location based on the building information marked in the building area drawing.

[0136] Step 430: For each of the one or more building area drawings, perform floor identification based on the building area drawings to determine auxiliary floor information.

[0137] Auxiliary floor information refers to floor information extracted from each building in one or more building area drawings.

[0138] In some embodiments, the processor 140 can extract the floor information marked in the drawings of each building area through image recognition technology (such as machine vision, OCR technology, etc.), and use the extracted floor information as auxiliary floor information.

[0139] In some embodiments, the processor 140 may also determine auxiliary floor information through a machine learning model.

[0140] In some embodiments, the processor 140 can input the building area drawing into the convolutional neural network to obtain a convolution feature vector; perform enhancement processing on the convolution feature vector to obtain an enhanced feature vector; and input the enhanced feature vector into the floor recognition model to determine auxiliary floor information.

[0141] In some embodiments, a convolutional neural network can be trained using a first sample and a first label. The first sample includes a large number of historical building area drawings, and the first label includes floor information from manually annotated historical building area drawings. After the convolutional neural network is trained, processor 140 can select a feature vector output by the previous layer of the convolutional neural network output layer as the convolutional feature vector.

[0142] In some embodiments, the processor 140 may also extract annotation information based on the building area drawings to obtain drawing annotation information; perform feature extraction on the drawing annotation information to obtain an annotation feature vector; and fuse the annotation feature vector and the convolution feature vector to obtain an enhanced feature vector.

[0143] Drawing marking information refers to the floor information marked in the building area drawings.

[0144] In some embodiments, the processor 140 can extract the floor information marked in the drawings of each building area through image recognition technology (such as machine vision, OCR technology, etc.), and use the extracted floor information as the drawing marking information.

[0145] In some embodiments, the processor 140 may further obtain a preset annotation area of ​​the target drawing and extract the drawing annotation information from the preset annotation area, wherein the preset annotation area is determined based on the annotation habits of the user who draws the target drawing.

[0146] The preset marking area refers to the area in the building area drawing that is preset for marking floor information.

[0147] Labeling habits refer to the user's habit of labeling spatial information such as buildings and floors when drawing drawings. For example, labeling habits may include labeling with conventional numbers (such as Arabic numerals and Roman numerals), labeling a single floor, labeling multiple floors uniformly, etc. In some embodiments, the user's labeling habits may be pre-stored in a storage device (such as storage device 150) in the form of vectors, and processor 140 may retrieve the labeling habits from the storage device based on the drawing user's name in the drawing label information.

[0148] In some embodiments of the present specification, a preset annotation area of ​​a target drawing is determined according to the annotation habits of the drawing user corresponding to the target drawing, and the drawing annotation information is extracted from the preset annotation area. This can better conform to the drawing habits of different users when extracting the annotation information, thereby improving the accuracy of the auxiliary floor information determined subsequently and improving the intelligence level of the target space information verification process.

[0149] In some embodiments, the processor 140 may further perform feature extraction on the drawing annotation information through methods such as text structure features, grammatical features, and word embedding to obtain an annotation feature vector.

[0150] In some embodiments, the processor 140 may fuse the labeled feature vector and the convolved feature vector in various ways. For example, the processor 140 may concatenate the labeled feature vector and the convolved feature vector to obtain an enhanced feature vector. In another example, the processor 140 may perform a weighted summation of the labeled feature vector and the convolved feature vector to obtain an enhanced feature vector, where the weights of the weighted summation may be preset.

[0151] In some embodiments, by fusing the annotation feature vector corresponding to the annotation information of the building area drawing with the convolution feature vector, an enhanced feature vector is obtained, which can be integrated into the annotation information of the drawing to improve the accuracy of the auxiliary floor information subsequently determined by the model.

[0152] The floor recognition model is a machine learning model used to determine auxiliary floor information in a building area drawing. In some embodiments, the floor recognition model can be a neural network model, a deep neural network model, etc.

[0153] In some embodiments, the input of the floor recognition model includes the enhanced feature vector, and the output includes the auxiliary floor information.

[0154] In some embodiments, the input to the floor recognition model also includes the target drawing's user's annotation habits and annotation parsing annotations. The annotation parsing annotations refer to an interpretation of the target drawing's user's annotation habits. For example, for the annotation "2nd floor," the annotation parsing annotations explain whether the user's annotation refers to "2nd floor" or "1st and 2nd floor."

[0155] In some embodiments of this specification, the input of the floor recognition model also includes the marking habits and marking analysis annotations of the user who draws the target drawing, which can improve the accuracy of the output auxiliary floor information and make the output results closer to the marking habits of different drawing users, making it more intelligent.

[0156] In some embodiments, a floor recognition model can be trained using a second sample and a second label. The second sample includes a large number of historical enhanced feature vectors corresponding to historical building area drawings, and the second label includes floor information from manually annotated historical building area drawings. When the large number of historical building area drawings in the second sample are consistent with the large number of historical building area drawings in the first sample, the first label and the second label may be the same.

[0157] In some embodiments, a convolutional neural network may be part of the floor recognition model.

[0158] In some embodiments of this specification, auxiliary floor information is determined through a machine learning model, which is more accurate and intelligent and can improve the verification accuracy and efficiency of the target space information.

[0159] Step 440: Combine the auxiliary building information and the auxiliary floor information to obtain auxiliary space information.

[0160] Auxiliary space information refers to the space information determined based on the building area drawings.

[0161] In some embodiments, the processor 140 may determine the auxiliary space information based on a method similar to that used to determine the target space information. For example, for each of one or more building area drawings, the processor 140 may combine the auxiliary building information and the auxiliary floor information in the building area drawing to obtain the auxiliary space information of the building area drawing. For example, the auxiliary floor information of a building area drawing includes "1st floor", "2nd floor", and "equipment floor", and the auxiliary building information includes "Building 1" and "Building 2". The processor 140 may combine each of the auxiliary floor information with each of the auxiliary building information, and then merge the combination results of the same building to obtain the auxiliary space information of the building area drawing: "Building 1: 1st floor, 2nd floor, equipment floor", "Building 2: 1st floor, 2nd floor, equipment floor".

[0162] In some embodiments, for each of one or more building area drawings, the processor 140 can respectively execute the method described in steps 420 to 440 (for example, simultaneously or separately) to obtain the auxiliary space information of each building area drawing, and then merge the auxiliary space information of each building area drawing as the auxiliary space information of the target drawing.

[0163] Step 450: Determine whether the auxiliary space information is consistent with the target space information.

[0164] In some embodiments, the processor 140 can determine whether the auxiliary spatial information and the target spatial information are consistent using various methods. For example, the processor 140 can directly compare the auxiliary spatial information and the target spatial information to determine whether they are consistent. In another example, the processor 140 can calculate the vector distance between the auxiliary spatial information and the target spatial information. When the vector distance is less than a distance threshold, it indicates that the auxiliary spatial information and the target spatial information are consistent; otherwise, they are inconsistent. Exemplary vector distances include one or more of Euclidean distance, cosine distance, Manhattan distance, and Chebyshev distance.

[0165] In some embodiments, when the auxiliary space information and the target space information are consistent, it indicates that the target space information is extracted correctly, and the processor 140 ends the process.

[0166] In some embodiments, when the auxiliary space information and the target space information are inconsistent, it indicates that there may be an error in the extraction of the target space information, and the processor 140 executes step 460 .

[0167] Step 460: In response to the inconsistency between the auxiliary space information and the target space information, output prompt information.

[0168] The prompt information is used to prompt the user to verify the target space information. For example, the prompt information may include a pop-up reminder, a highlighted reminder, etc.

[0169] In some embodiments, the prompt information also includes specific verification content, such as the specific building and floor of the target space information that is inconsistent with the auxiliary space information.

[0170] In some embodiments, the user can manually verify target space information that is inconsistent with auxiliary space information. In some embodiments, the processor 140 can also repeat one or more of processes 200, 300, and 400 to automatically verify the target space information. For example, the processor 140 can compare the target space information obtained after the repeated execution to see if it is consistent with the auxiliary space information or the target space information obtained before the repeated execution.

[0171] In some embodiments of the present specification, target spatial information is verified based on the target drawing, which can improve the recognition accuracy of the target spatial information. At the same time, when the extracted target spatial information is inconsistent with the auxiliary spatial information, a prompt can be promptly issued to the user for further verification, thereby further improving the recognition accuracy of the target spatial information. In addition, compared with the method of directly comparing with the spatial information in the spatial information database, the auxiliary spatial information of the target drawing is for the same drawing, the comparison object is more accurate, and the comparison result can be obtained quickly.

[0172] Figure 6 This is an exemplary module diagram of a drawing space information extraction system according to some embodiments of this specification. Figure 6 As shown, the drawing space information extraction system 600 includes a first acquisition module 610 , an identification module 620 , a comparison module 630 , a second acquisition module 640 and a third acquisition module 650 .

[0173] The first acquisition module 610 is configured to acquire a frame of a target drawing. In some embodiments, the frame includes label information.

[0174] The recognition module 620 is configured to perform image label recognition based on the image frame and determine initial spatial information.

[0175] The comparison module 630 is configured to compare the initial spatial information based on a preset spatial information database to obtain a comparison result.

[0176] The second acquisition module 640 is configured to split the initial space information according to the verification result to obtain space description information.

[0177] In some embodiments, the second acquisition module 640 is further configured to split the initial spatial information to obtain multiple spatial information combinations when the verification result meets the first preset condition; for each spatial information combination, obtain floor spatial description information and building spatial description information based on a preset method.

[0178] In some embodiments, the second acquisition module 640 is further configured to extract information related to floors based on the spatial information combination to obtain floor spatial description information; and extract information related to buildings based on the spatial information combination to obtain building spatial description information.

[0179] The third acquisition module 650 is configured to acquire target space information of the target drawing based on the space description information.

[0180] In some embodiments, the third acquisition module 650 is further configured to perform keyword extraction based on the floor space description information and the building space description information to obtain the target floor information and the target building information; combine the target floor information and the target building information to obtain the target space information of the target drawing.

[0181] In some embodiments, the third acquisition module 650 is further configured to extract the first number and the custom floor in the floor space description information; determine the first floor information based on the adjacent connecting words of the first number; determine the second floor information based on the consecutive words between the first number; determine the target floor information based on the first floor information and the second floor information, and use the custom floor as supplementary floor information; extract the second number and the custom building in the building space description information; determine the first building information based on the adjacent connecting words of the second number; determine the second building information based on the consecutive words between the second number; determine the target building information based on the first building information and the second building information, and use the custom building as supplementary building information.

[0182] In some embodiments, when the verification result meets the second preset condition, the third acquisition module 650 is further configured to split the initial space information to obtain floor space description information; perform keyword extraction based on the floor space description information to obtain target floor information; obtain sub-item information of the drawing frame, and based on the sub-item information, obtain sub-item space information; split the sub-item space information to obtain building space description information; perform keyword extraction based on the building space description information to obtain target building information; combine the target floor information and the target building information to obtain the target space information of the target drawing.

[0183] In some embodiments, the paper space information extraction system 600 further includes a verification module 660 .

[0184] In some embodiments, the verification module 660 is further configured to perform image segmentation based on the target drawing to obtain one or more building area drawings; determine auxiliary building information based on the one or more building area drawings; for each of the one or more building area drawings, perform floor identification based on the building area drawing to determine auxiliary floor information; combine the auxiliary building information and the auxiliary floor information to obtain auxiliary space information; determine whether the auxiliary space information and the target space information are consistent; in response to the inconsistency between the auxiliary space information and the target space information, output a prompt message; the prompt message is used to prompt the user to verify the target space information.

[0185] In some embodiments, the verification module 660 is further configured to input the building area drawing into the convolutional neural network to obtain a convolution feature vector; perform enhancement processing on the convolution feature vector to obtain an enhanced feature vector; and input the enhanced feature vector into the floor recognition model to determine the auxiliary floor information.

[0186] For a detailed description of the functions of each module of the drawing space information extraction system 600 shown above, please refer to the relevant parts of the previous text of this specification, for example, Figures 2 to 5 and related instructions.

[0187] It should be noted that the above description of the drawing space information extraction system 600 and its modules is only for convenience of description and does not limit this specification to the scope of the embodiments. Figure 6 The first acquisition module 610, identification module 620, comparison module 630, second acquisition module 640, third acquisition module 650, and verification module 660 disclosed in the specification may be different modules within a system, or a single module may implement the functions of two or more of the aforementioned modules. For example, the modules may share a storage module, or each module may have its own storage module. Such variations are within the scope of protection of this specification.

[0188] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example and does not constitute a limitation to this specification.

[0189] This specification also uses specific terms to describe the embodiments of this specification. Terms such as "one embodiment," "an embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that the mention of "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different places in this specification does not necessarily refer to the same embodiment.

[0190] It should be noted that, in order to simplify the description of this specification and facilitate understanding of one or more embodiments, the foregoing description of the embodiments of this specification sometimes combines multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not mean that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single embodiment disclosed above.

Claims

1. A drawing space information extraction method, characterized in that: The method comprises: Obtaining a drawing frame of a target drawing; the drawing frame includes drawing label information; Performing image label recognition based on the image frame to determine initial spatial information; Comparing the initial spatial information based on a preset spatial information database to obtain a comparison result; Splitting the initial spatial information according to the comparison result to obtain spatial description information; Based on the spatial description information, acquiring target spatial information of the target drawing; The step of splitting the initial spatial information according to the comparison result to obtain spatial description information includes: When the comparison result is a first comparison result, the initial spatial information is split to obtain a plurality of spatial information combinations, wherein the first comparison result is that the floor keyword and the building keyword in the initial spatial information are successfully matched with the floor keyword and the building keyword in the spatial information database; For each of the spatial information combinations, floor spatial description information and building spatial description information are obtained based on a preset method.

2. The method according to claim 1, wherein The acquiring target space information of the target drawing based on the space description information includes: Perform keyword extraction based on the floor space description information and the building space description information to obtain target floor information and target building information; The target floor information and the target building information are combined to obtain the target space information of the target drawing.

3. The method according to claim 1, wherein The obtaining of the floor space description information and the building space description information based on a preset method includes: Extracting information related to the floor based on the spatial information combination to obtain the floor space description information; Based on the spatial information combination, information related to the building is extracted to obtain the building space description information.

4. The method according to claim 2, wherein The keyword extraction based on the floor space description information and the building space description information to obtain target floor information and target building information includes: Extracting the first number and the custom floor in the floor space description information; determining first floor information based on adjacent connectives of the first digit; determining second floor information based on the consecutive words between the first digits; determining the target floor information based on the first floor information and the second floor information, and using the custom floor as supplementary floor information; Extracting the second number and the custom building from the building space description information; Determining first building information based on adjacent connectives of the second digit; Determining second building information based on the consecutive words between the second digits; Based on the first building information and the second building information, the target building information is determined, and the customized building is used as supplementary building information.

5. The method according to claim 1, wherein The method further comprises: When the comparison result is a second comparison result, the initial space information is split to obtain floor space description information, and the second comparison result is that the floor keyword in the initial space information successfully matches the floor keyword in the space information database; Perform keyword extraction based on the floor space description information to obtain target floor information; Acquire sub-item information of the frame, and acquire sub-item spatial information based on the sub-item information; Splitting the sub-item spatial information to obtain building spatial description information; Extract keywords based on the building space description information to obtain target building information; The target floor information and the target building information are combined to obtain the target space information of the target drawing.

6. The method according to claim 1, wherein The method further comprises: Perform image segmentation based on the target drawing to obtain one or more building area drawings; Determining auxiliary building information based on the one or more building area drawings; For each of the one or more building area drawings, performing floor identification based on the building area drawings to determine auxiliary floor information; Combining the auxiliary building information and the auxiliary floor information to obtain auxiliary space information; determining whether the auxiliary space information is consistent with the target space information; In response to the inconsistency between the auxiliary space information and the target space information, a prompt message is output; the prompt message is used to prompt a user to verify the target space information.

7. The method according to claim 6, wherein The performing floor identification based on the building area drawing and determining auxiliary floor information includes: Input the building area drawing into a convolutional neural network to obtain a convolution feature vector; Performing enhancement processing on the convolution feature vector to obtain an enhanced feature vector; The enhanced feature vector is input into a floor recognition model to determine the auxiliary floor information.

8. A drawing space information extraction system, characterized in that: The system comprises: A first acquisition module is configured to acquire a frame of a target drawing; the frame includes drawing label information; a recognition module configured to perform image label recognition based on the image frame and determine initial spatial information; a comparison module, configured to compare the initial spatial information based on a preset spatial information database to obtain a comparison result; A second acquisition module is configured to split the initial spatial information according to the comparison result to obtain spatial description information; a third acquisition module, configured to acquire target space information of the target drawing based on the space description information; The step of splitting the initial spatial information according to the comparison result to obtain spatial description information includes: When the comparison result is a first comparison result, the initial spatial information is split to obtain a plurality of spatial information combinations, wherein the first comparison result is that the floor keyword and the building keyword in the initial spatial information are successfully matched with the floor keyword and the building keyword in the spatial information database; For each of the spatial information combinations, floor spatial description information and building spatial description information are obtained based on a preset method.

9. A drawing space information extraction device, characterized in that: The apparatus includes at least one storage medium and at least one processor; The at least one storage medium is used to store computer instructions; The at least one processor is configured to execute the computer instructions to implement the method according to any one of claims 1 to 7.

Images