CN112927194A - Automatic checking method and system for design drawing and real object - Google Patents
Automatic checking method and system for design drawing and real object Download PDFInfo
- Publication number
- CN112927194A CN112927194A CN202110146190.8A CN202110146190A CN112927194A CN 112927194 A CN112927194 A CN 112927194A CN 202110146190 A CN202110146190 A CN 202110146190A CN 112927194 A CN112927194 A CN 112927194A
- Authority
- CN
- China
- Prior art keywords
- equipment
- neural network
- real object
- standard
- design drawing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013461 design Methods 0.000 title claims abstract description 118
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000013528 artificial neural network Methods 0.000 claims abstract description 100
- 238000013135 deep learning Methods 0.000 claims abstract description 49
- 238000012549 training Methods 0.000 claims description 36
- 238000003860 storage Methods 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 10
- 125000004122 cyclic group Chemical group 0.000 claims description 9
- 238000004826 seaming Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 abstract description 6
- 238000009826 distribution Methods 0.000 abstract description 6
- 238000007689 inspection Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 13
- 238000004590 computer program Methods 0.000 description 11
- 238000013473 artificial intelligence Methods 0.000 description 6
- 230000000306 recurrent effect Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000005096 rolling process Methods 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000013526 transfer learning Methods 0.000 description 3
- 238000012512 characterization method Methods 0.000 description 2
- 238000004821 distillation Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/42—Document-oriented image-based pattern recognition based on the type of document
- G06V30/422—Technical drawings; Geographical maps
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a method and a system for automatically checking a design drawing and a real object, wherein the method comprises the following steps: the method comprises the steps of identifying equipment line wiring data from a design drawing and an equipment real object by respectively adopting a deep learning neural network and a convolution/circulation neural network, automatically checking the line wiring data identified from the equipment real object and the standard line wiring data of the equipment by taking the equipment line wiring data in the design drawing as a standard, judging whether the wiring of the equipment real object is correct or not, automatically identifying the design drawing as unified equipment/line wiring data, realizing the quick and real-time checking of the field equipment real object wiring data, greatly reducing the workload of power distribution wiring/communication wiring management and greatly reducing manual inspection errors.
Description
Technical Field
The invention relates to the field of artificial intelligence, in particular to an automatic checking method and system for design drawings and real objects.
Background
The consistency of drawings and equipment wiring is often required to be checked in engineering implementation acceptance, operation, maintenance and overhaul, such as wiring in a distribution line control box or wiring of communication equipment.
Conventionally, whether a drawing is consistent with a field actual line or not is checked in a manual mode, the manual checking mode is low in efficiency and easy to make mistakes, and various potential safety hazards caused by human errors cannot be avoided even if various measures such as person responsibility, standardized naming/numbering, field photographing and the like are adopted in succession in recent years.
Disclosure of Invention
The present invention provides a method and system for automatically checking design drawings against physical objects that overcomes or at least partially solves the above-mentioned problems.
According to a first aspect of the present invention, there is provided a method for automatically checking a design drawing against a real object, comprising: inputting each electronic design drawing into a trained deep learning neural network, and acquiring an equipment number identified from each electronic equipment drawing by the deep learning neural network and standard line wiring data corresponding to the equipment number; storing each electronic design drawing, the identified corresponding equipment number and the standard circuit wiring data in a database; inputting the obtained image inside and outside the equipment real object into a trained convolution/circulation neural network, and obtaining an equipment number and corresponding line connection data which are identified by the convolution/circulation neural network according to the image inside and outside the equipment real object; according to the equipment number identified from the image inside and outside the equipment real object, standard circuit wiring data corresponding to the equipment number is searched from a database; and automatically checking the line connection data identified from the image inside and outside the equipment real object with the standard line connection data to determine whether the line connection data is correct.
On the basis of the technical scheme, the invention can be improved as follows.
Optionally, the inputting each electronic design drawing into the trained deep learning neural network, and the obtaining of the device number identified from each electronic device drawing by the deep learning neural network and the standard line connection data corresponding to the device number includes: inputting an electronic design drawing in a DWG format, a word format or a pdf format into a trained deep learning neural network, and identifying a seaming and standard construction design drawing from the electronic design drawing by the deep learning neural network; the volume seal comprises a project name, a design unit and effective time, and the standard construction design drawing comprises the project name, an equipment number and internal and external wiring data of the equipment.
Optionally, the internal and external wiring data of the device includes element numbers or number combinations inside the design chart, line numbers or number combinations between elements, and wiring relationships between elements.
Optionally, the deep learning neural network is trained by: acquiring a first training data set, wherein the first training data set comprises a plurality of standard design drawings, each standard design drawing is marked with an equipment number and standard circuit wiring data, and the plurality of standard design drawings comprise standard design drawings of different equipment manufacturers; and training the deep learning neural network by adopting the first training data set.
Optionally, the inputting the obtained inside and outside image of the device in real object into the trained convolution/circulation neural network, and the obtaining of the device number and the corresponding line connection data identified by the convolution/circulation neural network according to the inside and outside image of the device in real object includes: shooting and acquiring an equipment real object internal and external image sequence, wherein the equipment real object internal and external image sequence at least comprises a plurality of equipment outer box images and a plurality of equipment internal images; and inputting the shot image sequence inside and outside the equipment real object into the trained convolution/circulation neural network, identifying the equipment number by the convolution/circulation neural network according to the plurality of pieces of equipment outer box images, and identifying the equipment line wiring data according to the plurality of pieces of equipment inner images.
Optionally, the multiple in-device images include multi-angle in-device connection terminal and pipeline images, and the identified device line connection data includes a connection bank number, an access point sequence, a pipeline number, and a spatial correspondence relationship therebetween.
Optionally, the convolutional/recurrent neural network is trained by: acquiring a second training data set, wherein the second training data set comprises a plurality of equipment real object images, and each equipment real object image is marked with an equipment number and equipment line wiring data; training a convolutional/cyclic neural network using the second training data set.
According to a second aspect of the present invention, there is provided an automatic checking system for a design drawing and a material object, comprising: the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for inputting each electronic design drawing into a trained deep learning neural network and acquiring a device number identified from each electronic device drawing by the deep learning neural network and standard line wiring data corresponding to the device number; the storage module is used for storing each electronic design drawing, the identified corresponding equipment number and the standard circuit wiring data in a database; the second acquisition module is used for inputting the acquired images inside and outside the equipment real object into the trained convolution/circulation neural network and acquiring the equipment number and the corresponding line wiring data which are identified by the convolution/circulation neural network according to the images inside and outside the equipment real object; the searching module is used for searching standard circuit wiring data corresponding to the equipment number from a database according to the equipment number identified from the image inside and outside the equipment real object; and the checking module is used for automatically checking the line connection data identified from the images inside and outside the equipment real object with the standard line connection data so as to determine whether the line connection data is correct or not.
According to a third aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein the processor is configured to implement the steps of the method for automatically checking a design drawing against a real object when executing a computer management-like program stored in the memory.
According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having stored thereon a computer management-like program, which when executed by a processor, implements the steps of the method for automatically collating a design drawing with a physical object.
According to the automatic checking method and system for the design drawing and the real object, the deep learning neural network and the convolution/circulation neural network are respectively adopted to identify the equipment line wiring data from the design drawing and the equipment real object, the equipment line wiring data identified from the equipment real object and the standard line wiring data of the equipment are automatically checked by taking the equipment line wiring data in the design drawing as a standard, whether the wiring of the equipment real object is correct or not is judged, the design drawing can be automatically identified into the unified equipment/line wiring data, the quick and real-time checking of the field equipment real object wiring data can be realized, the workload of power distribution wiring/communication wiring management can be greatly reduced, and the man-made checking errors can be greatly reduced.
Drawings
FIG. 1 is a flow chart of an automatic checking method for design drawings and objects provided by the present invention;
FIG. 2 is a structural diagram of an automatic checking system for design drawings and objects according to the present invention;
FIG. 3 is a schematic diagram of a hardware structure of a possible electronic device provided in the present invention;
fig. 4 is a schematic diagram of a hardware structure of a possible computer-readable storage medium according to the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
Fig. 1 is a flowchart of an automatic checking method for a design drawing and a real object provided by the present invention, as shown in fig. 1, the method includes: 101. inputting each electronic design drawing into the trained deep learning neural network, and acquiring the equipment number identified from each electronic equipment drawing by the deep learning neural network and standard line wiring data corresponding to the equipment number; 102. storing each electronic design drawing, the identified corresponding equipment number and the standard circuit wiring data in a database; 103. inputting the obtained image inside and outside the equipment real object into the trained convolution/circulation neural network, and obtaining the equipment number and the corresponding line connection data which are identified by the convolution/circulation neural network according to the image inside and outside the equipment real object; 104. according to the equipment number identified from the image inside and outside the equipment real object, standard circuit wiring data corresponding to the equipment number is searched from a database; 105. and automatically checking the line connection data identified from the image inside and outside the equipment real object with the standard line connection data to determine whether the line connection data is correct.
It can be understood that, based on the conventional method of manually checking whether the line connection data of each device in the design drawing is matched with the line connection data of the real field device, whether the line connection of the real field device is correct is judged, and then the line connection data of the field device is checked. The manual checking mode is inefficient and is prone to errors.
In recent years, with the continuous deepening of the standardization of the engineering implementation process, the basic conditions of data matched and checked by drawing and a real object are gradually met, meanwhile, the artificial intelligence technology is gradually deeply applied to a plurality of fields such as images, texts, voice, unmanned driving and the like, the computing capacity of artificial intelligence chips (GPU and NPU) is also continuously improved, and the technical conditions for solving the power distribution wiring/communication wiring management problems through artificial intelligence are also mature.
Based on the method, the invention provides a method capable of automatically checking the line wiring data of each device in the design drawing and the line wiring data of the field device real object, and a Convolutional Neural Network (CNN) is the most mature image feature extraction technology at present and is the technical basis in the fields of automatic image classification, target detection/tracking and the like. The convolutional neural network can process character sequences in the images, has the learning capacity of characteristic map space transformation, and can be used for automatic deformation and restoration of the images. A Recurrent Neural Network (RNN) is a type of Recurrent Neural Network (Recurrent Neural Network) that takes sequence data as input and performs recursion (recursion) in the evolution direction of the sequence, and all nodes (Recurrent units) are connected in a chain manner, and can be used to extract feature data from sequences such as video and the like for classification or tracking.
The method for automatically checking the design drawings and the equipment real objects comprises the steps of inputting each electronic design drawing into a trained deep learning neural network, and identifying equipment numbers and line connection data of the equipment from each electronic design drawing through the deep learning neural network. The line connection data of the equipment in the design drawing is standard line connection data, and each equipment number identified from the design drawing and the corresponding standard line connection data are stored in a database.
And for the field equipment real object, shooting a plurality of internal and external images of the equipment real object, inputting the shot plurality of internal and external images of the equipment into the trained convolution/circulation neural network, and identifying the equipment number and the corresponding line connection data from the plurality of internal and external images of the equipment real object by the convolution/circulation neural network.
Because the line connection data of the field device real object needs to be checked whether to be correct, the corresponding standard line connection data is searched from the database according to the device number identified from the image inside and outside the field device real object. And comparing and matching the line wiring data corresponding to the equipment number identified from the images inside and outside the equipment real object with the line wiring data of the equipment identified from the design drawing so as to determine whether the field line wiring of the equipment real object is correct.
The invention respectively adopts the deep learning neural network and the convolution/circulation neural network to identify the equipment line wiring data from the design drawing and the equipment real object, automatically checks the line wiring data identified from the equipment real object with the standard equipment line wiring data in the design drawing as a standard, judges whether the wiring of the equipment real object is correct or not, can automatically identify the design drawing as the unified equipment/line wiring data, can realize the fast and real-time checking of the field equipment real object wiring data, can greatly reduce the workload of power distribution wiring/communication wiring management, and greatly reduce the man-made inspection error.
In a possible embodiment, inputting each electronic design drawing into a deep learning neural network after training, and acquiring a device number identified from each electronic device drawing by the deep learning neural network and standard line connection data corresponding to the device number includes: inputting various digitized image results obtained after scanning of an electronic design drawing or a paper drawing in a DWG format, a word format or a pdf format into a trained deep learning neural network, and identifying a seaming and standard construction design drawing from the electronic design drawing by the deep learning neural network; the rolling seal comprises a project name, a design unit and effective time, and the standard construction design drawing comprises the project name, an equipment number and internal and external wiring data of the equipment.
It can be understood that the power distribution wiring/communication wiring design drawings provided by different equipment suppliers have different standards, styles and sizes, the problems of seals, stains, folds, fuzziness and the like also exist after the drawings are scanned and digitized, the number of drawings is accumulated to be increased by geometric multiples every year, and the traditional manual processing mode cannot meet the actual needs.
The invention trains and optimizes the deep learning neural network by adopting an end-to-end model training method based on a small amount of supervised data, can effectively remove interference factors such as stains of design drawings, and greatly reduces labor cost and errors.
The method comprises the following steps that a design drawing of an electronic edition can be in a DWG format or a word format or a pdf format or various digitized image results obtained after paper drawings are scanned, the design drawing of the electronic edition or the digitized image of the drawing is input into a deep learning neural network after training, and a character part and a graph part in the design drawing are identified by the deep learning neural network, wherein necessary contents such as a rolling seal, an in-roll catalog and a construction design drawing in a standard design drawing can be identified, wherein the rolling seal comprises contents such as a project name, a design unit and effective time, and an equipment supplier can be obtained through identification of the rolling seal; the standard construction design drawing comprises contents such as project names, equipment names, internal/external wiring drawings and the like, and the equipment names or equipment numbers and corresponding line wiring data can be identified through identifying the standard equipment drawings. In general, the construction design drawings of the same equipment manufacturer have consistency in design style (such as equipment characterizations, pipeline characterizations, and other design element elements) within a certain period of time.
The deep learning neural network mainly comprises two parts in a design drawing, wherein one part is used for extracting contents such as equipment manufacturers, construction projects, time ranges and the like from an input design drawing file and is used as a drawing classification basis; secondly, the circuit wiring data of the equipment can be obtained by adopting a plug-and-play mode for assembly and taking charge of capturing the serial numbers or serial number combinations of the components in different design drawings/tables and the line serial numbers or serial number combinations and wiring relations among the components.
In one possible embodiment, the deep learning neural network is trained by: acquiring a first training data set, wherein the first training data set comprises a plurality of standard design drawings, each standard design drawing is marked with an equipment number and standard circuit wiring data, and the plurality of standard design drawings comprise standard design drawings of different equipment manufacturers; the deep learning neural network is trained using a first training data set.
It is understood that deep Learning neural networks are primarily trained using a "Transfer Learning" approach, aided by a "Distillation" training. The transfer learning mode is used for completing the training of the deep learning neural network on the basis of a large number of standardized design drawings, and in practical use, the optimization training is performed aiming at standardized design elements such as element symbols, tables, connecting lines and the like or design drawings of the same equipment manufacturer at different periods. Distillation training is mainly used for rapidly building a model aiming at application side requirements.
The large number of standard design drawings comprise standard design drawings of different equipment suppliers, wherein for each standard design drawing, equipment numbers and line wiring data in the standard design drawing are marked. The deep learning neural network is trained by adopting a first training data set which comprises a plurality of standard design drawings and is marked with equipment numbers and line connection data, and the trained deep learning neural network can identify the equipment numbers and the line connection data in the standard design drawings. When identifying the line wiring data of the equipment, the method mainly captures the component numbers or number combinations in different design drawings/tables and the line numbers or number combinations and wiring relations among the components.
In a possible embodiment, inputting the obtained real-object internal and external images of the device into a trained convolutional/cyclic neural network, and obtaining the device number and the corresponding line connection data identified by the convolutional/cyclic neural network according to the real-object internal and external images of the device includes: shooting and acquiring an equipment real object internal and external image sequence, wherein the equipment real object internal and external image sequence at least comprises a plurality of equipment outer box images and a plurality of equipment internal images; and inputting the shot image sequence inside and outside the equipment real object into the trained convolution/circulation neural network, identifying the equipment number by the convolution/circulation neural network according to the plurality of pieces of equipment outer box images, and identifying the equipment line wiring data according to the plurality of pieces of equipment inner images.
It can be understood that, for the field device real object which is already wired, images or videos of a plurality of device real objects are shot to form an image sequence or a video sequence, wherein the image sequence or the video sequence comprises an outer box image of the device and a wiring terminal and a pipeline image in the device, and for the outer box of the device and the inside of the device, shooting can be carried out at a plurality of angles, and multi-angle tracking shooting is carried out inside and outside the device. And inputting the shot image sequence inside and outside the equipment real object into the trained convolution/circulation neural network, and extracting the equipment number on the equipment outer box from the image of the equipment outer box by the convolution/circulation neural network. And identifying the serial numbers and the position relations of the access pipelines and the corresponding terminals thereof from the images of the connecting terminals and the pipelines in the equipment, namely identifying the serial number and the serial number of a certain terminal accessed by the pipeline and the serial number of the corresponding terminal, identifying the serial number, the access point sequence, the serial number and the space corresponding relation of the pipelines in the equipment, and obtaining the line wiring data of the equipment through the identified information such as the serial number, the access point sequence, the serial number and the space corresponding relation of the pipelines.
In one possible embodiment, the convolutional/cyclic neural network is trained by: acquiring a second training data set, wherein the second training data set comprises a plurality of equipment real object images, and each equipment real object image is marked with an equipment number and equipment line wiring data; training a convolutional/cyclic neural network using the second training data set.
It is understood that for convolutional/recurrent neural networks, training is done in a "Transfer Learning" manner. Specifically, internal and external images of various field device real objects are collected, device numbers and device line connection data in the images are marked to form a second training data set for training a convolution/circulation neural network, the convolution/circulation neural network is trained by the second training data set, and the trained convolution/circulation neural network is used for identifying the device numbers and the line connection data in the internal and external images of the device real objects.
Identifying equipment numbers and standard line wiring data from each standard design drawing by adopting a deep learning neural network, and correspondingly storing each standard equipment drawing and the equipment numbers and the line wiring data identified from the standard equipment drawing in a database; and identifying the equipment number and the equipment line wiring data of the field equipment from the internal and external images of the field equipment real object by adopting a convolution/circulation neural network. According to the identified equipment number of the field equipment, standard line wiring data corresponding to the field equipment is searched from a database, actual line wiring data of the field equipment is compared and matched with the standard line wiring data, and when the actual line wiring data of the field equipment is consistent with the standard line wiring data, the line wiring data of the field equipment is correct; if not, it indicates that the line connection data of the field device is incorrect and needs to be readjusted.
The invention adopts a deep learning neural network to automatically identify and digitally process key elements (type, number, wiring structure, wiring point and external circuit) of various types of design drawings to form a searchable drawing database; according to an image/image sequence or video mode, different real objects and incidental contents (such as symbols, characters and the like) thereof are automatically identified on site by adopting a convolution/circulation neural network, and whether the elements such as the positions, labels, pipelines, access/connection points and the like of the real objects are consistent with the design of a drawing is quickly checked through drawing data.
And (4) finishing the matching check of the drawing and the real object during wiring overhaul by using an image sequence or a video. For example, in a screen cabinet of a certain power enterprise device, a camera of a mobile device is used for shooting a real object, and the shooting mode of a power line control box body is to shoot an image of the real object to be detected from top to bottom. And inputting the multi-frame sequential images into a convolution/circulation neural network for identification, so that whether the real object wiring conforms to the drawing can be accurately identified.
The invention realizes drawing identification, field object identification and drawing and object consistency check based on artificial intelligence on the basis of a small amount of training data, greatly improves the working efficiency and correctness of equipment during acceptance, operation and maintenance, and takes a screen cabinet with 50 interfaces as an example, the manual check needs about 40 minutes, and the machine check only needs 30 seconds. The method adopts an image sequence or video mode to realize the characteristic identification of a real object box body, a line board, a pipeline, an access point and the like based on artificial intelligence, can realize quick retrieval and reading of drawing data, needs at least 5 minutes for manually searching dozens of pages and hundreds of pages of equipment drawings in the prior art, and has machine retrieval within 1 second.
Fig. 2 is a structural diagram of an automatic checking system of a design drawing and an actual object according to an embodiment of the present invention, and as shown in fig. 2, the automatic checking system of a design drawing and an actual object includes a first obtaining module 201, a storage module 202, a second obtaining module 203, a searching module 204, and a checking module 205, where: a first obtaining module 201, configured to input each electronic design drawing into a trained deep learning neural network, and obtain an equipment number identified from each electronic equipment drawing by the deep learning neural network and standard wiring line data corresponding to the equipment number; the storage module 202 is used for storing each electronic design drawing, the identified corresponding equipment number and the standard wiring line data in a database; a second obtaining module 203, configured to input the obtained inside and outside images of the device real object into the trained convolution/circulation neural network, and obtain a device number and corresponding connection line data that are identified by the convolution/circulation neural network according to the inside and outside images of the device real object; the searching module 204 is used for searching standard wiring line data corresponding to the equipment number from a database according to the equipment number identified from the image inside and outside the equipment real object; and the checking module 204 is configured to automatically check the wiring line data identified from the inside and outside real object images of the device with the standard wiring line data to determine whether the wiring line data is correct.
It should be understood that the automatic checking system for design drawings and real objects provided in this embodiment corresponds to the automatic checking method for design drawings and real objects provided in the foregoing embodiments, and the technical features of the automatic checking system for design drawings and real objects may refer to the technical features of the automatic checking method for design drawings and real objects of the foregoing embodiments, and will not be described repeatedly herein.
Referring to fig. 3, fig. 3 is a schematic view of an embodiment of an electronic device according to an embodiment of the present disclosure. As shown in fig. 3, an electronic device according to an embodiment of the present application includes a memory 310, a processor 320, and a computer program 311 stored in the memory 320 and executable on the processor 320, where the processor 5320 executes the computer program 311 to implement the following steps: inputting each electronic design drawing into the trained deep learning neural network, and acquiring the equipment number identified from each electronic equipment drawing by the deep learning neural network and standard wiring line data corresponding to the equipment number; storing each electronic design drawing, the identified corresponding equipment number and the standard wiring line data in a database; inputting the obtained image inside and outside the equipment real object into the trained convolution/circulation neural network, and obtaining the equipment number and the corresponding wiring line data which are identified by the convolution/circulation neural network according to the image inside and outside the equipment real object; according to the equipment number identified from the image inside and outside the equipment real object, searching standard wiring line data corresponding to the equipment number from a database; and automatically checking the wiring line data identified from the images inside and outside the equipment object with the standard wiring line data to determine whether the wiring line data is correct.
Referring to fig. 4, fig. 4 is a schematic diagram illustrating an embodiment of a computer-readable storage medium according to the present application. As shown in fig. 4, the present embodiment provides a computer-readable storage medium 400, on which a computer program 411 is stored, the computer program 411 implementing the following steps when executed by a processor: inputting each electronic design drawing into the trained deep learning neural network, and acquiring the equipment number identified from each electronic equipment drawing by the deep learning neural network and standard wiring line data corresponding to the equipment number; storing each electronic design drawing, the identified corresponding equipment number and the standard wiring line data in a database; inputting the obtained image inside and outside the equipment real object into the trained convolution/circulation neural network, and obtaining the equipment number and the corresponding wiring line data which are identified by the convolution/circulation neural network according to the image inside and outside the equipment real object; according to the equipment number identified from the image inside and outside the equipment real object, searching standard wiring line data corresponding to the equipment number from a database; and automatically checking the wiring line data identified from the images inside and outside the equipment object with the standard wiring line data to determine whether the wiring line data is correct.
It should be noted that, in the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to relevant descriptions of other embodiments for parts that are not described in detail in a certain embodiment.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include such modifications and variations.
Claims (10)
1. An automatic checking method for a design drawing and a real object is characterized by comprising the following steps:
inputting each electronic design drawing into a trained deep learning neural network, and acquiring an equipment number identified from each electronic equipment drawing by the deep learning neural network and standard line wiring data corresponding to the equipment number;
storing each electronic design drawing, the identified corresponding equipment number and the standard circuit wiring data in a database;
inputting the obtained image inside and outside the equipment real object into a trained convolution/circulation neural network, and obtaining an equipment number and corresponding line connection data which are identified by the convolution/circulation neural network according to the image inside and outside the equipment real object;
according to the equipment number identified from the image inside and outside the equipment real object, standard circuit wiring data corresponding to the equipment number is searched from a database;
and automatically checking the line connection data identified from the image inside and outside the equipment real object with the standard line connection data to determine whether the line connection data is correct.
2. The method according to claim 1, wherein the step of inputting each electronic design drawing into the trained deep learning neural network, and the step of obtaining the device number identified from each electronic device drawing by the deep learning neural network and the standard line connection data corresponding to the device number includes:
inputting an electronic design drawing in a DWG format, a word format or a pdf format into a trained deep learning neural network, and identifying a seaming and standard construction design drawing from the electronic design drawing by the deep learning neural network;
the volume seal comprises a project name, a design unit and effective time, and the standard construction design drawing comprises the project name, an equipment number and internal and external wiring data of the equipment.
3. The method of claim 2, wherein the internal and external wiring data of the device includes a component number or a combination of numbers in the design drawing, a line number or a combination of numbers between components, and a wiring relationship between components.
4. The method for automatically collating design drawings and physical objects according to any one of claims 1 to 3, wherein the deep learning neural network is trained by:
acquiring a first training data set, wherein the first training data set comprises a plurality of standard design drawings, each standard design drawing is marked with an equipment number and standard circuit wiring data, and the plurality of standard design drawings comprise standard design drawings of different equipment manufacturers;
and training the deep learning neural network by adopting the first training data set.
5. The method according to claim 1, wherein the step of inputting the obtained images of the inside and outside of the device in the real object into a trained convolutional/cyclic neural network, and the step of obtaining the device number and the corresponding line connection data identified by the convolutional/cyclic neural network according to the images of the inside and outside of the device in the real object comprises:
shooting and acquiring an equipment real object internal and external image sequence, wherein the equipment real object internal and external image sequence at least comprises a plurality of equipment outer box images and a plurality of equipment internal images;
and inputting the shot image sequence inside and outside the equipment real object into the trained convolution/circulation neural network, identifying the equipment number by the convolution/circulation neural network according to the plurality of pieces of equipment outer box images, and identifying the equipment line wiring data according to the plurality of pieces of equipment inner images.
6. The method according to claim 5, wherein the plurality of in-device images include multi-angle in-device wiring terminal and pipeline images, and the identified device line wiring data includes a wiring bar number, an access point sequence, a pipeline number, and a spatial correspondence thereof.
7. The method of automatically collating design drawings and physical objects according to claim 1, 5 or 6, wherein the convolutional/cyclic neural network is trained by:
acquiring a second training data set, wherein the second training data set comprises a plurality of equipment real object images, and each equipment real object image is marked with an equipment number and equipment line wiring data;
training a convolutional/cyclic neural network using the second training data set.
8. An automatic checking system for design drawings and real objects is characterized by comprising:
the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for inputting each electronic design drawing into a trained deep learning neural network and acquiring a device number identified from each electronic device drawing by the deep learning neural network and standard line wiring data corresponding to the device number;
the storage module is used for storing each electronic design drawing, the identified corresponding equipment number and the standard circuit wiring data in a database;
the second acquisition module is used for inputting the acquired images inside and outside the equipment real object into the trained convolution/circulation neural network and acquiring the equipment number and the corresponding line wiring data which are identified by the convolution/circulation neural network according to the images inside and outside the equipment real object;
the searching module is used for searching standard circuit wiring data corresponding to the equipment number from a database according to the equipment number identified from the image inside and outside the equipment real object;
and the checking module is used for automatically checking the line connection data identified from the images inside and outside the equipment real object with the standard line connection data so as to determine whether the line connection data is correct or not.
9. An electronic device, comprising a memory and a processor, wherein the processor is configured to implement the steps of the method for automatically checking a design drawing with a physical object according to any one of claims 1 to 7 when executing a computer management program stored in the memory.
10. A computer-readable storage medium, on which a computer management-like program is stored, which when executed by a processor, implements the steps of the method for automatically checking a design drawing against an actual object according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110146190.8A CN112927194A (en) | 2021-02-02 | 2021-02-02 | Automatic checking method and system for design drawing and real object |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110146190.8A CN112927194A (en) | 2021-02-02 | 2021-02-02 | Automatic checking method and system for design drawing and real object |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112927194A true CN112927194A (en) | 2021-06-08 |
Family
ID=76167852
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110146190.8A Pending CN112927194A (en) | 2021-02-02 | 2021-02-02 | Automatic checking method and system for design drawing and real object |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112927194A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113822288A (en) * | 2021-11-24 | 2021-12-21 | 广东电网有限责任公司湛江供电局 | Method and system for quickly checking white heads of secondary line cables |
CN114494681A (en) * | 2022-01-24 | 2022-05-13 | 江门市电力工程输变电有限公司 | Intelligent comparison method and system for wiring terminal, computer equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2014181966A (en) * | 2013-03-18 | 2014-09-29 | Nuflare Technology Inc | Inspection method and inspection apparatus |
CN111242024A (en) * | 2020-01-11 | 2020-06-05 | 北京中科辅龙科技股份有限公司 | Method and system for recognizing legends and characters in drawings based on machine learning |
CN111563509A (en) * | 2020-04-30 | 2020-08-21 | 贵州电网有限责任公司 | Tesseract-based substation terminal row identification method and system |
CN111860525A (en) * | 2020-08-06 | 2020-10-30 | 宁夏宁电电力设计有限公司 | Bottom-up optical character recognition method suitable for terminal block |
-
2021
- 2021-02-02 CN CN202110146190.8A patent/CN112927194A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2014181966A (en) * | 2013-03-18 | 2014-09-29 | Nuflare Technology Inc | Inspection method and inspection apparatus |
CN111242024A (en) * | 2020-01-11 | 2020-06-05 | 北京中科辅龙科技股份有限公司 | Method and system for recognizing legends and characters in drawings based on machine learning |
CN111563509A (en) * | 2020-04-30 | 2020-08-21 | 贵州电网有限责任公司 | Tesseract-based substation terminal row identification method and system |
CN111860525A (en) * | 2020-08-06 | 2020-10-30 | 宁夏宁电电力设计有限公司 | Bottom-up optical character recognition method suitable for terminal block |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113822288A (en) * | 2021-11-24 | 2021-12-21 | 广东电网有限责任公司湛江供电局 | Method and system for quickly checking white heads of secondary line cables |
CN114494681A (en) * | 2022-01-24 | 2022-05-13 | 江门市电力工程输变电有限公司 | Intelligent comparison method and system for wiring terminal, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110992317B (en) | PCB defect detection method based on semantic segmentation | |
US11763443B2 (en) | Method for monitoring manufacture of assembly units | |
WO2020177432A1 (en) | Multi-tag object detection method and system based on target detection network, and apparatuses | |
CN110348441B (en) | Value-added tax invoice identification method and device, computer equipment and storage medium | |
CN103093208B (en) | A kind of method and system of fruits and vegetables identification | |
CN112927194A (en) | Automatic checking method and system for design drawing and real object | |
CN109446689A (en) | DC converter station electrical secondary system drawing recognition methods and system | |
CN110533654A (en) | The method for detecting abnormality and device of components | |
CN110032654B (en) | Supermarket commodity entry method and system based on artificial intelligence | |
CN109284700B (en) | Method, storage medium, device and system for detecting multiple faces in image | |
CN110610169A (en) | Picture labeling method and device, storage medium and electronic device | |
CN114049568A (en) | Object shape change detection method, device, equipment and medium based on image comparison | |
CN113821674A (en) | Intelligent cargo supervision method and system based on twin neural network | |
CN114168768A (en) | Image retrieval method and related equipment | |
CN112621765B (en) | Automatic equipment assembly control method and device based on manipulator | |
CN111626313B (en) | Feature extraction model training method, image processing method and device | |
Abbas | Recovering homography from camera captured documents using convolutional neural networks | |
CN111968103A (en) | Steel coil spacing detection method, system, medium and electronic terminal | |
CN111382710A (en) | Drawing comparison method based on image recognition | |
CN111898555A (en) | Book checking identification method, device, equipment and system based on images and texts | |
CN114155471A (en) | Design drawing and object verification method, device, computer equipment and system | |
CN112733948B (en) | Full-target associated index automatic testing method, system, medium and terminal | |
CN114227717A (en) | Intelligent inspection method, device, equipment and storage medium based on inspection robot | |
CN111783786A (en) | Picture identification method and system, electronic equipment and storage medium | |
CN112131418A (en) | Target labeling method, target labeling device and computer-readable storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210608 |