CN116756835A - Template combination design method, device, equipment and storage medium - Google Patents
Template combination design method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116756835A CN116756835A CN202311023748.9A CN202311023748A CN116756835A CN 116756835 A CN116756835 A CN 116756835A CN 202311023748 A CN202311023748 A CN 202311023748A CN 116756835 A CN116756835 A CN 116756835A
- Authority
- CN
- China
- Prior art keywords
- data
- template
- wall
- wall body
- feature
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000013461 design Methods 0.000 title claims abstract description 49
- 238000003860 storage Methods 0.000 title claims abstract description 19
- 238000010276 construction Methods 0.000 claims abstract description 75
- 230000003044 adaptive effect Effects 0.000 claims abstract description 63
- 230000008569 process Effects 0.000 claims abstract description 19
- 230000000007 visual effect Effects 0.000 claims abstract description 15
- 239000013598 vector Substances 0.000 claims description 84
- 238000012545 processing Methods 0.000 claims description 29
- 230000007246 mechanism Effects 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 16
- 238000007781 pre-processing Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 13
- 230000004044 response Effects 0.000 claims description 12
- 238000009827 uniform distribution Methods 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 10
- 238000007499 fusion processing Methods 0.000 claims description 10
- 238000013135 deep learning Methods 0.000 claims description 9
- 238000009499 grossing Methods 0.000 claims description 8
- 238000013210 evaluation model Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000000547 structure data Methods 0.000 claims description 3
- 230000009897 systematic effect Effects 0.000 abstract description 2
- 230000007613 environmental effect Effects 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 238000005457 optimization Methods 0.000 description 6
- 238000005286 illumination Methods 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000002787 reinforcement Effects 0.000 description 2
- 229910000838 Al alloy Inorganic materials 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003042 antagnostic effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000011449 brick Substances 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000003562 lightweight material Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000003973 paint Substances 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 239000002023 wood Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- 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
-
- 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
- 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
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- 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
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/806—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of extracted features
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Geometry (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Computational Mathematics (AREA)
- Structural Engineering (AREA)
- Civil Engineering (AREA)
- Mathematical Optimization (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Architecture (AREA)
- Computer Graphics (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention relates to the technical field of template application, and discloses a template combination design method, a device, equipment and a storage medium. The template combination design method comprises the following steps: collecting first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body; acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body; the invention not only greatly improves the accuracy and efficiency of selecting the construction template, but also reduces the human error in the manual selection process, so that the construction process is more scientific, systematic and intelligent. By fusing various types of characteristic data, the actual condition of the wall can be more comprehensively and accurately understood and considered, and therefore, a more adaptive construction template is recommended.
Description
Technical Field
The present invention relates to the field of template application technologies, and in particular, to a template combination design method, device, equipment, and storage medium.
Background
The aluminum alloy building templates are popularized in the global scope in recent years, and have been widely applied in the house building industry. In order to reduce the production cost and improve the universality, the existing building templates mainly adopt standard templates, and in the building industry, the wall construction is an important and complex link, and the related factors are very many.
Conventional construction methods often require manual inspection of the wall to determine the best construction method and form to use, which is inefficient and may be subject to errors. The development of big data and machine learning techniques offers new possibilities for automation and accuracy, but these techniques need to be fully exploited by appropriate models and data acquisition means.
Disclosure of Invention
The invention provides a template combination design method, a template combination design device, template combination design equipment and a storage medium, which are used for improving the accuracy and efficiency of selecting construction templates.
The first aspect of the present invention provides a template combination design method, which includes:
collecting first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
Acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
extracting features of the first feature data to obtain a first wall feature vector, extracting features of the second feature data to obtain a second wall feature vector, and performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
inputting the target feature vector and the use data of the historical construction template into a trained self-adaptive template library model for intelligent recommendation to obtain a self-adaptive template; the self-adaptive template library model is trained in advance;
acquiring a plurality of parameters of an adaptive template, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
assigning a priority to each adaptive template type, and determining template construction parameters according to parameters of the adaptive templates and the priorities; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
Optionally, in a first implementation manner of the first aspect of the present invention, the collecting, by a preset vision sensor, first feature data of the wall includes:
acquiring minimum diameter distances from the visual sensor to all positions of the wall body, converting all minimum diameter distance data into point cloud data, wherein each point cloud data represents space coordinates corresponding to each set element in a minimum diameter distance set, and the combination of all space coordinates is used for representing the three-dimensional shape of the wall body;
converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm; the three-dimensional model is used for representing the shape, the size, the position and the structure data of the wall body;
collecting image data of a wall body, and preprocessing the image data; wherein the pretreatment comprises gray conversion and denoising treatment;
dividing the preprocessed image data based on a preset image dividing algorithm to obtain a wall area image;
extracting texture features of the wall area image through a preset HOG algorithm to obtain wall texture features;
calculating a smoothing factor and a roughness factor of the wall according to the obtained wall texture characteristics; the smoothness factor represents that the uniform distribution variance of the wall texture features is larger than a preset uniform distribution threshold of the wall texture features, and the roughness factor represents that the uniform distribution variance of the wall texture features is smaller than the preset uniform distribution threshold of the wall texture features;
Modeling the surface of the wall body according to the calculated smooth factors and rough factors to obtain a wall body surface evaluation model; the parameters of the wall surface evaluation model are used for reflecting the actual surface condition of the wall.
Optionally, in a second implementation manner of the first aspect of the present invention, the converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm includes:
scanning the wall body through a preset laser radar sensor to obtain three-dimensional point cloud data of the wall body;
preprocessing the three-dimensional point cloud data, wherein the preprocessing comprises noise filtering and data standardization;
performing marginalization processing on the preprocessed three-dimensional point cloud data based on a preset edge recognition algorithm, and recognizing and extracting a point set of the wall body through the distribution of the space points;
based on a preset feature matching algorithm, determining a three-dimensional plane equation corresponding to the point set of the wall body, and constructing a preliminary three-dimensional wall body model according to the three-dimensional plane equation;
and carrying out smoothing and texture mapping on the preliminary wall body three-dimensional model to obtain a final wall body three-dimensional model.
Optionally, in a third implementation manner of the first aspect of the present invention, the performing feature extraction on the first feature data to obtain a first wall feature vector, performing feature extraction on the second feature data to obtain a second wall feature vector, and performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector, where the performing step includes:
Setting first parameters of the Gabor filter, wherein the parameters comprise first direction parameters and first scale parameters of the Gabor filter;
decomposing the first characteristic data through a Gabor filter with a first parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a first wall characteristic vector;
setting second parameters of the Gabor filter, wherein the parameters comprise second direction parameters and second scale parameters of the Gabor filter;
decomposing the second characteristic data through a Gabor filter with a second parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a second wall characteristic vector;
and weighting the first wall body feature vector and the second wall body feature vector based on the first weight of the first wall body feature vector and the second weight of the second wall body feature vector to obtain a target feature vector.
Optionally, in a fourth implementation manner of the first aspect of the present invention, the training process of the adaptive template library model includes:
performing first processing on the target feature vectors in the training set and the use data of the historical construction templates through a combined attention mechanism to obtain first data after preliminary processing; wherein the combined attention mechanism includes a multi-head self-attention mechanism, a location self-attention mechanism, and a cross-modality self-attention mechanism;
Processing the preliminarily processed first data and the expected template using samples through a hierarchical global attention mechanism to generate a preliminary feature weight map; the preliminary feature weight map reflects the preliminary importance distribution of each feature to the target task;
clustering and dimension reduction processing is carried out on the first data after preliminary processing through a preset mechanism to obtain second data, each feature of the second data is extracted and learned through a deep learning processing algorithm, and a deep feature understanding map is generated; the depth feature understanding map reflects the key importance distribution of each feature to the target task;
training an initial prediction model by combining the initial feature weight map, the depth feature understanding map and an expected template using sample to obtain a self-adaptive template library model; the self-adaptive template library model is used for converting the use data of the target feature vector and the historical construction template into the self-adaptive template with personalized recommendation.
A second aspect of the present invention provides a form set designing apparatus including:
the first acquisition module is used for acquiring first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
The second acquisition module is used for acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
the fusion module is used for carrying out feature extraction on the first feature data to obtain a first wall feature vector, carrying out feature extraction on the second feature data to obtain a second wall feature vector, and carrying out data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
the recommendation module is used for inputting the target feature vector and the use data of the historical construction template into the trained self-adaptive template library model to conduct intelligent recommendation so as to obtain the self-adaptive template; the self-adaptive template library model is trained in advance;
the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring a plurality of parameters of an adaptive template, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
the processing module is used for distributing priority to each adaptive template type and determining template construction parameters according to the parameters of the adaptive templates and the priority; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
A third aspect of the present invention provides a template combination design apparatus, comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the template combination design apparatus to perform the template combination design method described above.
A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored therein that, when executed on a computer, cause the computer to perform the above-described template combination design method.
In the technical scheme provided by the invention, the beneficial effects are as follows: according to the template combination design method, the device, the equipment and the storage medium, a series of characteristic data of a wall body, including three-dimensional data, structural data, surface data, position data, lighting data, gap data and the like, are accurately acquired through the preset visual sensor and the high-definition camera, and the target characteristic vector is obtained through characteristic extraction and data fusion processing. And then, carrying out intelligent recommendation by using the trained self-adaptive template library model and carrying use data of the historical construction templates to obtain the self-adaptive templates, and finally determining template construction parameters according to parameters of the self-adaptive templates and the assigned priorities. The invention not only greatly improves the accuracy and efficiency of selecting the construction template, but also reduces the human error in the manual selection process, so that the construction process is more scientific, systematic and intelligent. By fusing various types of characteristic data, the actual condition of the wall can be more comprehensively and accurately understood and considered, and therefore, a more adaptive construction template is recommended. Meanwhile, through priority allocation and parameter adjustment of the template types, various requirements of construction can be met more flexibly and pertinently.
Drawings
FIG. 1 is a schematic diagram of one embodiment of a modular design method in accordance with an embodiment of the present invention;
FIG. 2 is a schematic diagram of an embodiment of a modular design apparatus according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a template combination design method, a device, equipment and a storage medium. The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For ease of understanding, a specific flow of an embodiment of the present invention is described below with reference to fig. 1, and an embodiment of a method for designing a module combination in an embodiment of the present invention includes:
step 101, acquiring first characteristic data of a wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
it will be appreciated that the execution body of the present invention may be a template combination design apparatus, and may also be a terminal or a server, which is not limited herein. The embodiment of the invention is described by taking a server as an execution main body as an example.
Specifically, in engineering construction, the collected first characteristic data of the wall body can provide basic and visual information about the wall body. This step is achieved by a preset vision sensor, which may be an optical scanner, a laser rangefinder or other device capable of acquiring 3D information.
In this context, they are each able to acquire the following data:
three-dimensional data: such data may be acquired by laser scanning or stereoscopic vision equipment to generate a three-dimensional model of the wall. The model can describe the shape, size and position of the wall in detail, and the information is key to selecting and designing a construction template.
For example, if three-dimensional data of a wall is obtained by a laser scanning device, the height, width and thickness of the wall can be clearly known, and the device can also recognize whether the wall is vertical or whether the wall is inclined to some extent.
Structural data: visual sensors such as near infrared scanners or high resolution cameras can provide structural information of the wall, such as material composition, cracks, voids, etc. This information is very helpful in understanding the strength, durability, and potential problems of the wall.
For example, a high resolution camera or near infrared scanner may detect the presence of cracks, holes, or signs of aging in the wall, thereby reflecting the actual condition of the wall.
Surface data: raw surface data may be obtained by ultra-high resolution cameras or laser scanning devices, including color, texture, gloss, etc. Such information can be used to determine the appearance and texture of the wall, as well as whether special handling is required.
For example, the roughness of a wall can be known by scanning the surface of the wall, and thus the difficulty of future construction thereon, or the presence of stains on the surface can be predicted, and the degree of cleaning required can be known.
The integration of these data can greatly improve the accuracy and efficiency of the engineering construction, providing key inputs for the next steps.
102, acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
specifically, in this step, a preset high-definition camera is used to collect second characteristic data of the wall. This high-precision data acquisition helps to understand the environmental conditions and microscopic features of the wall more deeply, which is critical to the selection and design of templates that best accommodate construction requirements.
Position data: first, the positional data of the wall body is of great importance for building understanding and positioning. The high definition camera can capture accurate information of the wall location, including the relative positions of other building elements (e.g., doors, windows, posts, etc.), and its positioning in the overall layout of the building. This information plays an important role in optimizing the construction plan and in allocating resources.
For example, the captured wall position data may be accurate relative distance of the wall to the window, or suggest that a floor should be constructed first or last in order to avoid affecting the construction.
Lighting data: and secondly, the high-definition camera can acquire lighting data of the wall body by recording the illumination condition of the surface of the wall body. The data can be used for evaluating the natural illumination environment of the wall body, and showing whether a shielding object, illumination intensity, angle and the like exist.
For example, if a camera tests that a wall area is in shadow for a substantial portion of the day, additional lighting may be required to ensure quality of construction when constructing the wall.
Gap data: finally, the gaps between the wall and other objects are observed through the data acquired by the high-definition camera. Such gap data may help understand the degree of physical isolation between walls and may even predict future sound and thermal insulation requirements.
For example, taking a photograph with a camera may find that there are irregular gaps between the wall and the ceiling, which would need to be taken into account in the design of the form.
Step 103, performing feature extraction on the first feature data to obtain a first wall feature vector, performing feature extraction on the second feature data to obtain a second wall feature vector, and performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
Specifically, for example, three-dimensional data of a wall obtained by a visual sensor shows the inclination of a certain section of the wall, and then the inclination is a feature to be extracted. For another example, if the illumination data collected by the high-definition camera shows that some walls are in shadows, "shadows" or "low illumination" are also a feature that needs to be extracted.
These features are then dimension-reduced using some specific algorithm, such as Principal Component Analysis (PCA), and the two feature data are combined to generate a composite target feature vector. The target feature vector is actually a data list with various features, and can comprehensively represent various attributes of the wall body.
104, inputting the target feature vector and the use data of the historical construction template into a trained self-adaptive template library model for intelligent recommendation to obtain a self-adaptive template; the self-adaptive template library model is trained in advance;
specifically, in this step, an adaptive template can be obtained by inputting the target feature vector and the usage data of the history construction template to the pre-trained adaptive template library model. The intelligent construction template recommendation method is based on a machine learning method and is used for intelligently recommending the construction template.
The self-adaptive template library model is a pre-trained machine learning model, and is trained by feeding a large amount of related data, including the use condition of a past construction template, various characteristic data of a wall body and the like. The model learns how to intelligently recommend the optimal construction templates based on different input data (i.e., the target feature vectors mentioned above).
For example, if the target feature vector shows that a wall is inclined and is in shadow most of the time, the model may recommend a template that is suitable for slope construction and for use in low light environments. The use data of the historical construction templates help the model to know which construction templates have the best effect under similar environment and wall characteristics.
Emphasis is placed on training and recommendation strategies for models. For example, deep learning, such as Deep Neural Networks (DNNs), may be used for model training, which may help mine more complex, non-linear relationships, making template recommendations more accurate. In the recommended strategy, a multi-objective optimization algorithm can be designed, so that the effect of a construction template is focused, and factors such as construction cost and time are focused to realize more comprehensive optimization.
Step 105, obtaining a plurality of parameters of an adaptive template, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
specifically, a plurality of parameters of the adaptive template are obtained, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types; please describe this step in depth and explain the refinement, by way of example, some inventive content related to this step is added to explain the refinement.
At this step, a plurality of parameters of the adaptive template are obtained. These parameters describe the characteristics of the respective adaptive template types and the corresponding design parameters.
Each adaptive form type may be designed specifically for a particular environment or construction task, such as for high humidity environments, for tilting walls, and the like. At the same time, each template type has a set of design parameters associated with it, including but not limited to size, material, shape, color, etc., which have a direct impact on construction efficiency and quality.
For example, a form type for a tilting wall may have specific design parameters, such as some adjustment to the tilt, and the material may be selected from lightweight materials that are easy to work on the tilting surface. While another template for low light environments may set the color parameters to a brighter color family to enhance visibility during construction.
The innovative content can be to develop a set of more elegant adaptive template parameter optimization system, and the system can dynamically adjust the template parameters according to real-time environmental factors and construction requirements. For example, by combining a deep learning algorithm with historical data and feedback, each template type parameter is self-corrected and optimized, so that the template type parameter has better adaptability to various environmental conditions and project requirements.
In addition, advanced artificial intelligence techniques, such as genetic algorithms, are introduced to optimize the template design parameters. After each construction is completed, the model can adjust parameters according to construction results to generate a new template design, and then the new template design is used in the next construction. Such iterative optimization processes can make adaptive template systems more and more intelligent to accommodate more scenarios and challenges.
Step 106, assigning a priority to each adaptive template type, and determining template construction parameters according to parameters of the adaptive templates and the priorities; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
Specifically, in this step, each adaptive template will be assigned a priority according to its characteristics and project requirements, and template construction parameters will be determined according to the parameters of the template and the priorities. These parameters will then be used to adjust the adaptive template to the target template to be used in the actual construction.
The setting of the priority of the adaptive template is a process determined according to various factors such as project requirements, environmental conditions, construction efficiency and the like. For example, if the project needs to be completed quickly within a defined time, templates with efficient construction characteristics will be given higher priority. Under special environmental conditions, such as low light and high humidity, templates with corresponding adaptability are given higher priority.
For example, assuming that one adaptive template is specifically designed for tilting walls and the other for use in a darkened environment, both templates would be given a higher priority when the sample wall is tilted and in a darkened condition. By combining the parameters of the two high-priority templates, a new template construction parameter suitable for the specific environment can be prepared.
The focus is on the process of using optimization algorithms to automate the determination of priorities and construction parameters. An algorithm can be created to automatically adjust priorities based on existing data and environmental requirements and to determine optimal construction parameters by data driven methods, saving labor and improving accuracy. In addition, the construction parameter determination process can be continuously optimized by means of a machine learning technology, so that the template decision process is more intelligent.
In order to better track and optimize the construction process, a feedback loop can be introduced into the process, namely, after each construction is integrally completed, the construction data are collected and analyzed, and the template library and the priority thereof are adjusted to perform continuous optimization. Thus, a self-learning and self-optimizing construction system is realized.
Another embodiment of the module combination design method in the embodiment of the invention comprises the following steps:
the first characteristic data of the wall body is collected through a preset visual sensor, and the method comprises the following steps:
acquiring minimum diameter distances from the visual sensor to all positions of the wall body, converting all minimum diameter distance data into point cloud data, wherein each point cloud data represents space coordinates corresponding to each set element in a minimum diameter distance set, and the combination of all space coordinates is used for representing the three-dimensional shape of the wall body;
Converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm; the three-dimensional model is used for representing the shape, the size, the position and the structure data of the wall body;
collecting image data of a wall body, and preprocessing the image data; wherein the pretreatment comprises gray conversion and denoising treatment;
dividing the preprocessed image data based on a preset image dividing algorithm to obtain a wall area image;
extracting texture features of the wall area image through a preset HOG algorithm to obtain wall texture features;
calculating a smoothing factor and a roughness factor of the wall according to the obtained wall texture characteristics; the smoothness factor represents that the uniform distribution variance of the wall texture features is larger than a preset uniform distribution threshold of the wall texture features, and the roughness factor represents that the uniform distribution variance of the wall texture features is smaller than the preset uniform distribution threshold of the wall texture features;
modeling the surface of the wall body according to the calculated smooth factors and rough factors to obtain a wall body surface evaluation model; the parameters of the wall surface evaluation model are used for reflecting the actual surface condition of the wall.
Specifically, in this step, the characteristic data of the wall is collected mainly through a preset vision sensor, which depends on a plurality of processes including space measurement, data conversion, image processing, texture feature capture, and evaluation of the surface properties of the wall.
First, the minimum diameter distance from each location to the wall is collected by a vision sensor. The data are converted into point cloud data, namely, each minimum diameter distance data are converted into corresponding space coordinates, and the three-dimensional shape of the wall body is formed by combining all the space coordinates. For example, laser radar technology may be used for measurement and data acquisition.
And then, converting the point cloud data into a more detailed and accurate three-dimensional model by using a preset three-dimensional reconstruction algorithm, wherein the shape, the size, the position and the structure of the wall body can be comprehensively described.
Then, the vision sensor collects the image data of the wall, and the data are preprocessed, such as gray level conversion and denoising, so that the vision characteristics of the image and the original wall are consistent, and the image is matched for subsequent processing steps.
The preprocessed image data is segmented through a preset image segmentation algorithm, and a wall area image is extracted. This is important because texture feature extraction and analysis can only be accurately performed if the wall area is correctly identified.
After the wall area image is identified, the texture feature extraction is carried out on the image in the wall area by using a preset HOG (Histogram of Oriented Gradients) algorithm, so that the wall texture feature is obtained.
Then, the smoothness factor and the roughness factor of the wall are calculated according to the extracted texture feature data, and can be used for evaluating the smoothness or the roughness of the surface of the wall.
And finally, modeling the wall surface according to the calculated smooth factors and the roughness factors to obtain a wall surface evaluation model, wherein parameters of the model reflect the actual conditions of the wall surface.
For AI-enhanced vision sensors, a Time of Flight (ToF) camera may be used that relies on measuring the Time of Flight of light to calculate object distance, enabling accurate measurement of the distance at each point, providing accurate three-dimensional information. In addition, an open source framework such as OpenCV or TensorFlow can be combined to construct a deep learning model for image data so as to improve the recognition and processing capacity of images.
For example, an unmanned aerial vehicle using a ToF camera and a deep learning algorithm can collect three-dimensional features of a wall during flight, and not only detect the shape and position of the wall, but also identify the material and color of the wall. After preprocessing the image, the model can automatically distinguish shadows, light temperatures, and other factors that may interfere with image analysis, distinguishing the wall from the background.
Image preprocessing and texture feature extraction algorithms can also be improved using deep learning techniques. For example, a pre-trained deep neural network such as AlexNet, VGG or ResNet may be used, and these models have been trained on millions of images, and have been learned to identify many common texture and image features. With this knowledge, the network can be fine-tuned to be able to identify and process different wall textures, such as brick, paint, wood, etc.
The wall surface assessment model may be built on top of reinforcement learning techniques. During training, the model is set to achieve as high accuracy as possible on preset indicators, such as "roughness" and "smoothness". With a large number of training sample inputs, the model will continually try to adjust its parameters to perform better on this task. Reinforcement learning cases such as AlphaGo have demonstrated that such approaches can reach and even exceed the performance of human experts.
For example, an AI model may be trained to score wall "smoothness" by viewing wall photographs of different textures and colors. After thousands of iterations, the model will eventually automatically identify and quantify the smoothness of various walls and can be used to predict wall texture and smoothness in unseen photographs.
Another embodiment of the module combination design method in the embodiment of the invention comprises the following steps:
the converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm comprises the following steps:
scanning the wall body through a preset laser radar sensor to obtain three-dimensional point cloud data of the wall body;
preprocessing the three-dimensional point cloud data, wherein the preprocessing comprises noise filtering and data standardization;
performing marginalization processing on the preprocessed three-dimensional point cloud data based on a preset edge recognition algorithm, and recognizing and extracting a point set of the wall body through the distribution of the space points;
based on a preset feature matching algorithm, determining a three-dimensional plane equation corresponding to the point set of the wall body, and constructing a preliminary three-dimensional wall body model according to the three-dimensional plane equation;
and carrying out smoothing and texture mapping on the preliminary wall body three-dimensional model to obtain a final wall body three-dimensional model.
Specifically, a preset radar sensor scans a wall: at this stage, the laser radar loaded by the device, for example, the laser radar of industrial grade such as Velodyne or Quanergy is used for scanning, so as to obtain three-dimensional point cloud data of the wall body. This step may be performed at an initial stage of construction or at a repair stage to obtain an accurate physical structure of the wall.
Preprocessing three-dimensional point cloud data: preprocessing will typically involve two steps, denoising and data normalization. Errors and noise in the scanned data are first removed using, for example, the RANSAC algorithm. And then the data are standardized by a statistical method, so that all data are in a common measurement space, and the subsequent processing is facilitated.
Edge recognition algorithm: here, a point set corresponding to the wall may be determined from the distribution of spatial points using a gradient-based edge detection means such as Sobel, prewitt, canny or the like. The innovation of the step is that the wall edges of different materials and different environments can be adaptively identified through an AI algorithm, which is difficult to realize in the traditional algorithm.
Feature matching: after identifying the point set of the wall, a plane fit is required using, for example, a random sample consensus algorithm (Random Sample Consensus, RANSAC). This is an iterative algorithm that selects the plane equation that best represents most of the data over multiple iterations. Also, deep learning can be used to improve this process as well, and planes can be identified faster and more accurately.
Constructing and optimizing a three-dimensional model: and constructing a preliminary three-dimensional model according to the plane equation obtained in the previous step. After the model is subjected to the steps of removing noise points, filling missing parts, executing smoothing processing such as a Laplacian smoothing algorithm and the like, a final three-dimensional model can be obtained. Texture mapping can then also be applied, which covers the model with the material and color information of the real wall, improving the fidelity of the model. In this step, a deep learning algorithm such as GAN (generation of an antagonistic network) may be used to generate a texture closer to the real world.
Another embodiment of the module combination design method in the embodiment of the invention comprises the following steps:
the step of extracting the first feature data to obtain a first wall feature vector, the step of extracting the second feature data to obtain a second wall feature vector, and the step of performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector comprises the following steps:
setting first parameters of the Gabor filter, wherein the parameters comprise first direction parameters and first scale parameters of the Gabor filter;
decomposing the first characteristic data through a Gabor filter with a first parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a first wall characteristic vector;
setting second parameters of the Gabor filter, wherein the parameters comprise second direction parameters and second scale parameters of the Gabor filter;
decomposing the second characteristic data through a Gabor filter with a second parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a second wall characteristic vector;
and weighting the first wall body feature vector and the second wall body feature vector based on the first weight of the first wall body feature vector and the second weight of the second wall body feature vector to obtain a target feature vector.
Another embodiment of the module combination design method in the embodiment of the invention comprises the following steps:
the training process of the self-adaptive template library model comprises the following steps:
performing first processing on the target feature vectors in the training set and the use data of the historical construction templates through a combined attention mechanism to obtain first data after preliminary processing; wherein the combined attention mechanism includes a multi-head self-attention mechanism, a location self-attention mechanism, and a cross-modality self-attention mechanism;
processing the preliminarily processed first data and the expected template using samples through a hierarchical global attention mechanism to generate a preliminary feature weight map; the preliminary feature weight map reflects the preliminary importance distribution of each feature to the target task;
clustering and dimension reduction processing is carried out on the first data after preliminary processing through a preset mechanism to obtain second data, each feature of the second data is extracted and learned through a deep learning processing algorithm, and a deep feature understanding map is generated; the depth feature understanding map reflects the key importance distribution of each feature to the target task;
training an initial prediction model by combining the initial feature weight map, the depth feature understanding map and an expected template using sample to obtain a self-adaptive template library model; the self-adaptive template library model is used for converting the use data of the target feature vector and the historical construction template into the self-adaptive template with personalized recommendation.
The method for designing a module assembly according to the embodiment of the present invention is described above, and the device for designing a module assembly according to the embodiment of the present invention is described below, referring to fig. 2, one embodiment of the device 1 for designing a module assembly according to the embodiment of the present invention includes:
the first acquisition module 11 is used for acquiring first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
the second acquisition module 12 is used for acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
the fusion module 13 is configured to perform feature extraction on the first feature data to obtain a first wall feature vector, perform feature extraction on the second feature data to obtain a second wall feature vector, and perform data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
the recommendation module 14 is configured to input the target feature vector and the usage data of the historical construction template to the trained adaptive template library model for intelligent recommendation, so as to obtain an adaptive template; the self-adaptive template library model is trained in advance;
An obtaining module 15, configured to obtain a plurality of parameters of an adaptive template, where the plurality of parameters of the adaptive template include a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
a processing module 16, configured to assign a priority to each adaptive template type, and determine a template construction parameter according to a parameter of the adaptive template and the priority; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
The invention also provides a template combination design device, which comprises a memory and a processor, wherein the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, cause the processor to execute the steps of the template combination design method in the above embodiments.
The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, or may be a volatile computer readable storage medium, in which instructions are stored which, when executed on a computer, cause the computer to perform the steps of the template combination design method.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (8)
1. The template combination design method is characterized by comprising the following steps of:
collecting first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
extracting features of the first feature data to obtain a first wall feature vector, extracting features of the second feature data to obtain a second wall feature vector, and performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
Inputting the target feature vector and the use data of the historical construction template into a trained self-adaptive template library model for intelligent recommendation to obtain a self-adaptive template; the self-adaptive template library model is trained in advance;
acquiring a plurality of parameters of an adaptive template, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
assigning a priority to each adaptive template type, and determining template construction parameters according to parameters of the adaptive templates and the priorities; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
2. The method of claim 1, wherein the acquiring, by a preset vision sensor, first feature data of the wall body comprises:
acquiring minimum diameter distances from the visual sensor to all positions of the wall body, converting all minimum diameter distance data into point cloud data, wherein each point cloud data represents space coordinates corresponding to each set element in a minimum diameter distance set, and the combination of all space coordinates is used for representing the three-dimensional shape of the wall body;
Converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm; the three-dimensional model is used for representing the shape, the size, the position and the structure data of the wall body;
collecting image data of a wall body, and preprocessing the image data; wherein the pretreatment comprises gray conversion and denoising treatment;
dividing the preprocessed image data based on a preset image dividing algorithm to obtain a wall area image;
extracting texture features of the wall area image through a preset HOG algorithm to obtain wall texture features;
calculating a smoothing factor and a roughness factor of the wall according to the obtained wall texture characteristics; the smoothness factor represents that the uniform distribution variance of the wall texture features is larger than a preset uniform distribution threshold of the wall texture features, and the roughness factor represents that the uniform distribution variance of the wall texture features is smaller than the preset uniform distribution threshold of the wall texture features;
modeling the surface of the wall body according to the calculated smooth factors and rough factors to obtain a wall body surface evaluation model; the parameters of the wall surface evaluation model are used for reflecting the actual surface condition of the wall.
3. The template combination design method according to claim 2, wherein the converting the point cloud data into a three-dimensional model based on a preset three-dimensional reconstruction algorithm includes:
scanning the wall body through a preset laser radar sensor to obtain three-dimensional point cloud data of the wall body;
preprocessing the three-dimensional point cloud data, wherein the preprocessing comprises noise filtering and data standardization;
performing marginalization processing on the preprocessed three-dimensional point cloud data based on a preset edge recognition algorithm, and recognizing and extracting a point set of the wall body through the distribution of the space points;
based on a preset feature matching algorithm, determining a three-dimensional plane equation corresponding to the point set of the wall body, and constructing a preliminary three-dimensional wall body model according to the three-dimensional plane equation;
and carrying out smoothing and texture mapping on the preliminary wall body three-dimensional model to obtain a final wall body three-dimensional model.
4. The method of claim 1, wherein the performing feature extraction on the first feature data to obtain a first wall feature vector, performing feature extraction on the second feature data to obtain a second wall feature vector, and performing data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector, includes:
Setting first parameters of the Gabor filter, wherein the parameters comprise first direction parameters and first scale parameters of the Gabor filter;
decomposing the first characteristic data through a Gabor filter with a first parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a first wall characteristic vector;
setting second parameters of the Gabor filter, wherein the parameters comprise second direction parameters and second scale parameters of the Gabor filter;
decomposing the second characteristic data through a Gabor filter with a second parameter set to obtain a plurality of filter responses, and carrying out characteristic statistics on each filter response to obtain a second wall characteristic vector;
and weighting the first wall body feature vector and the second wall body feature vector based on the first weight of the first wall body feature vector and the second weight of the second wall body feature vector to obtain a target feature vector.
5. The template assembly design method according to claim 1, wherein the training process of the adaptive template library model comprises:
performing first processing on the target feature vectors in the training set and the use data of the historical construction templates through a combined attention mechanism to obtain first data after preliminary processing; wherein the combined attention mechanism includes a multi-head self-attention mechanism, a location self-attention mechanism, and a cross-modality self-attention mechanism;
Processing the preliminarily processed first data and the expected template using samples through a hierarchical global attention mechanism to generate a preliminary feature weight map; the preliminary feature weight map reflects the preliminary importance distribution of each feature to the target task;
clustering and dimension reduction processing is carried out on the first data after preliminary processing through a preset mechanism to obtain second data, each feature of the second data is extracted and learned through a deep learning processing algorithm, and a deep feature understanding map is generated; the depth feature understanding map reflects the key importance distribution of each feature to the target task;
training an initial prediction model by combining the initial feature weight map, the depth feature understanding map and an expected template using sample to obtain a self-adaptive template library model; the self-adaptive template library model is used for converting the use data of the target feature vector and the historical construction template into the self-adaptive template with personalized recommendation.
6. A template assembly design apparatus, comprising:
the first acquisition module is used for acquiring first characteristic data of the wall body through a preset visual sensor; the first characteristic data comprise three-dimensional data of a wall body, structural data of the wall body and surface data of the wall body;
The second acquisition module is used for acquiring second characteristic data of the wall body through a preset high-definition camera; the second characteristic data comprise position data of the wall body, lighting data of the wall body and gap data of the wall body;
the fusion module is used for carrying out feature extraction on the first feature data to obtain a first wall feature vector, carrying out feature extraction on the second feature data to obtain a second wall feature vector, and carrying out data fusion processing on the first wall feature vector and the second wall feature vector to obtain a target feature vector;
the recommendation module is used for inputting the target feature vector and the use data of the historical construction template into the trained self-adaptive template library model to conduct intelligent recommendation so as to obtain the self-adaptive template; the self-adaptive template library model is trained in advance;
the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring a plurality of parameters of an adaptive template, wherein the plurality of parameters of the adaptive template comprise a plurality of adaptive template types and template design parameters corresponding to the adaptive template types;
the processing module is used for distributing priority to each adaptive template type and determining template construction parameters according to the parameters of the adaptive templates and the priority; and adjusting the adaptive template according to the template construction parameters to obtain a target template.
7. A template combination design apparatus, characterized in that the template combination design apparatus comprises: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the template combination design apparatus to perform the template combination design method of any one of claims 1-5.
8. A computer-readable storage medium having instructions stored thereon, which when executed by a processor, implement the template combination design method of any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311023748.9A CN116756835B (en) | 2023-08-15 | 2023-08-15 | Template combination design method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311023748.9A CN116756835B (en) | 2023-08-15 | 2023-08-15 | Template combination design method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116756835A true CN116756835A (en) | 2023-09-15 |
| CN116756835B CN116756835B (en) | 2023-12-01 |
Family
ID=87948120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311023748.9A Active CN116756835B (en) | 2023-08-15 | 2023-08-15 | Template combination design method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116756835B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012034236A1 (en) * | 2010-09-16 | 2012-03-22 | Ambercore Software Inc. | System and method for detailed automated feature extraction from data having spatial coordinates |
| US20180232900A1 (en) * | 2016-04-08 | 2018-08-16 | Orbital Insight, Inc. | Remote determination of containers in geographical region |
| CN110222609A (en) * | 2019-05-24 | 2019-09-10 | 江西理工大学 | A kind of wall body slit intelligent identification Method based on image procossing |
| CN110287962A (en) * | 2019-05-20 | 2019-09-27 | 平安科技（深圳）有限公司 | Remote Sensing Target extracting method, device and medium based on superobject information |
| CN114066891A (en) * | 2022-01-17 | 2022-02-18 | 北京市建设工程质量第一检测所有限责任公司 | Wall crack identification and judgment method, device, equipment and medium based on deep learning |
| CN114662208A (en) * | 2022-05-19 | 2022-06-24 | 安徽智享云科技有限公司 | Construction visualization system and method based on Bim technology |
| CN115240040A (en) * | 2022-06-07 | 2022-10-25 | 北京理工大学 | Method and device for enhancing human behavior characteristics of through-wall radar |
| CN115758487A (en) * | 2022-07-15 | 2023-03-07 | 广东三维家信息科技有限公司 | Household template matching method and device, electronic equipment and storage medium |
| KR20230108103A (en) * | 2022-01-10 | 2023-07-18 | 인하대학교 산학협력단 | Apparatus for checking the construction level of the window-to-wall joint using thermal image and fast numerical model and method thereof |
-
2023
- 2023-08-15 CN CN202311023748.9A patent/CN116756835B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012034236A1 (en) * | 2010-09-16 | 2012-03-22 | Ambercore Software Inc. | System and method for detailed automated feature extraction from data having spatial coordinates |
| US20180232900A1 (en) * | 2016-04-08 | 2018-08-16 | Orbital Insight, Inc. | Remote determination of containers in geographical region |
| CN110287962A (en) * | 2019-05-20 | 2019-09-27 | 平安科技（深圳）有限公司 | Remote Sensing Target extracting method, device and medium based on superobject information |
| CN110222609A (en) * | 2019-05-24 | 2019-09-10 | 江西理工大学 | A kind of wall body slit intelligent identification Method based on image procossing |
| KR20230108103A (en) * | 2022-01-10 | 2023-07-18 | 인하대학교 산학협력단 | Apparatus for checking the construction level of the window-to-wall joint using thermal image and fast numerical model and method thereof |
| CN114066891A (en) * | 2022-01-17 | 2022-02-18 | 北京市建设工程质量第一检测所有限责任公司 | Wall crack identification and judgment method, device, equipment and medium based on deep learning |
| CN114662208A (en) * | 2022-05-19 | 2022-06-24 | 安徽智享云科技有限公司 | Construction visualization system and method based on Bim technology |
| CN115240040A (en) * | 2022-06-07 | 2022-10-25 | 北京理工大学 | Method and device for enhancing human behavior characteristics of through-wall radar |
| CN115758487A (en) * | 2022-07-15 | 2023-03-07 | 广东三维家信息科技有限公司 | Household template matching method and device, electronic equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 姚海云;舒红;汪善华;曾坤;: "多定位源数据分析的特征向量空间滤波模型及仿真实验", 城市勘测, no. 02 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116756835B (en) | 2023-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111008597B (en) | Space identification method and device for CAD drawing, electronic equipment and storage medium | |
| CN112785643A (en) | Indoor wall corner two-dimensional semantic map construction method based on robot platform | |
| US11982628B2 (en) | System and method for detecting defects on imaged items | |
| KR102254773B1 (en) | Automatic decision and classification system for each defects of building components using image information, and method for the same | |
| CN111563442A (en) | Slam method and system for fusing point cloud and camera image data based on laser radar | |
| EP3340106A1 (en) | Method for assigning particular classes of interest within measurement data | |
| CN111862119A (en) | Semantic information extraction method based on Mask-RCNN | |
| CN110929696A (en) | Remote sensing image semantic segmentation method based on multi-mode attention and self-adaptive fusion | |
| CN110348368B (en) | Method, computer readable medium and system for artificial intelligence analysis of house type graph | |
| JP6781432B2 (en) | Radio wave propagation simulation model creation method, creation system, creation device and creation program | |
| CN112330593A (en) | Building surface crack detection method based on deep learning network | |
| Laga et al. | Image-based plant stornata phenotyping | |
| CN111523540A (en) | Metal surface defect detection method based on deep learning | |
| CN117011477B (en) | BIM-based steel structure deformation monitoring and processing method and system | |
| Lukyanov et al. | Optical Method For Monitoring Tool Control For Green Burnishing With Using Of Algorithms With Adaptive Settings | |
| CN114548253A (en) | Digital twin model construction system based on image recognition and dynamic matching | |
| CN116756835B (en) | Template combination design method, device, equipment and storage medium | |
| CN117152094A (en) | Method, device and system for analyzing surface defects of steel plate based on computer vision | |
| CN116777865A (en) | Underwater crack identification method, system, device and storage medium | |
| Dulecha et al. | Crack detection in single-and multi-light images of painted surfaces using convolutional neural networks | |
| CN111275020A (en) | Room state identification method | |
| CN113284158B (en) | Image edge extraction method and system based on structural constraint clustering | |
| Sapkota | Segmentation of coloured point cloud data | |
| Anitha et al. | A survey on crack detection algorithms for concrete structures | |
| CN114926675A (en) | Method and device for detecting shell stain defect, computer equipment and 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |