CN111192363B - User power distribution room design generation method based on cloud computing - Google Patents
User power distribution room design generation method based on cloud computing Download PDFInfo
- Publication number
- CN111192363B CN111192363B CN201911341997.6A CN201911341997A CN111192363B CN 111192363 B CN111192363 B CN 111192363B CN 201911341997 A CN201911341997 A CN 201911341997A CN 111192363 B CN111192363 B CN 111192363B
- Authority
- CN
- China
- Prior art keywords
- neural network
- convolutional neural
- power distribution
- depth
- distribution room
- 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.)
- Active
Links
Classifications
-
- 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
- G06T17/05—Geographic models
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Remote Sensing (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Human Resources & Organizations (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a user power distribution room design generation method based on cloud computing, and belongs to the technical field of industrial expansion engineering. The existing industrial expansion design method has the defects that more time is required for completing design work and efficiency is low. The three-dimensional information and pictures of the power distribution room are collected through the laser depth sensor data, parameters required by drawing are obtained through a convolutional neural network, and the parameters are transmitted to a CAD module; finally outputting a practically available distribution room design drawing; the industrial expansion design time can be effectively shortened, and the industrial expansion design efficiency is improved. The invention can shorten the design period, standardize engineering standard, increase supply and expense, promote user satisfaction and finally continuously promote the generation of enterprise value by organically integrating power supply service, power engineering design, power engineering construction, power equipment manufacture and the like, thereby forming a business mode of virtuous circle development.
Description
Technical Field
The invention relates to a user power distribution room design generation method based on cloud computing, and belongs to the technical field of industrial expansion engineering.
Background
The high-voltage expansion engineering is generally divided into three types of temporary expansion, high-voltage new installation and Gao Yakuo capacity expansion, wherein the high-voltage expansion requires on-site collection of high-voltage distribution equipment data installed at a user side, and the current working flow is that after a customer manager has a good requirement, a designer completes on-site checking work to collect data of a user power distribution room and equipment, and a photographing and power distribution room measurement sketch drawing method is mainly adopted. And after the on-site funding is finished, redrawing the drawn sketch in CAD software, and then combining the equipment design to be installed with the existing funding design according to the expansion requirements of users to form a new distribution room design drawing. After the drawing is confirmed by a client manager and a user, a designer draws a construction drawing again and gives the construction drawing to an implementation person to finish equipment installation work.
The existing industrial expansion design method needs about 1 week to finish the design work under all smooth conditions. This process consumes too much time for the grid, design, construction, and the customer themselves, so new technology needs to be introduced to shorten the workflow.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a method for collecting three-dimensional information and pictures of a power distribution room through laser depth sensor data, acquiring parameters required by drawing by utilizing a convolutional neural network and transmitting the parameters to a CAD module; finally, the actually available distribution room design drawing is output, the business expansion design time can be effectively shortened, and the business expansion design efficiency is improved.
In order to achieve the above purpose, the technical scheme of the invention is as follows:
a user power distribution room design generation method based on cloud computing,
which comprises the following steps:
firstly, establishing an interface program WSGI of a Web server, a Web application program and a depth computing framework which are defined by using a voice for python programming;
secondly, collecting three-dimensional information of the power distribution room by using a laser depth camera through staff on the scene of the power distribution room;
thirdly, transmitting the three-dimensional reconstruction information and the laser depth photo of the equipment to a WSGI of a cloud server through a wireless network;
transmitting data to a convolutional neural network reasoning reference module based on the GPU cluster through a WSGI interface of a cloud, acquiring parameters required by drawing, and transmitting the parameters to a pythonAPI interface of a CAD module;
and fifthly, using a cloud acceleration computing architecture to accelerate the space computation amount of data collected by the site sensor, and finally outputting a standard drawing of the site fund collection of the power distribution room.
The three-dimensional information and pictures of the power distribution room are collected through the laser depth sensor data, parameters required by drawing are obtained through a convolutional neural network, and the parameters are transmitted to a CAD module; finally outputting a practically available distribution room design drawing; the industrial expansion design time can be effectively shortened, and the industrial expansion design efficiency is improved.
By applying the scheme of the invention, the existing high-voltage industrial expansion design flow can be reduced from the original 5 working days to 0.5 working days, and the revolutionary improvement of the working efficiency is realized. The invention can shorten the design period, standardize engineering standard, increase supply and expense, promote user satisfaction and finally continuously promote the generation of enterprise value by organically integrating power supply service, power engineering design, power engineering construction, power equipment manufacture and the like, thereby forming a business mode of virtuous circle development.
As a preferred technical measure:
the input data of the convolutional neural network is a three-dimensional reconstruction point cloud of a power distribution room, which is collected by a laser depth camera through a synchronous positioning and mapping technology SLAM, a partial differential algorithm is applied to the point cloud, points with large variation are obtained to serve as characteristic point clouds, and then a plurality of coordinate points are randomly extracted to serve as an input source of the convolutional neural network; the device image taken by the laser depth camera is used as another input to the convolutional neural network.
The three-dimensional data collection is realized through synchronous positioning and mapping technology, and the method can be combined with on-site three-dimensional data collection, and can be independently applied, so that the data collection efficiency can be effectively improved, and meanwhile, the diversity and the accuracy of the input data of the convolutional neural network are improved.
As a preferred technical measure:
training the convolutional neural network, completing data collection through a simulation system, and randomly generating a plurality of power distribution rooms and equipment with different sizes by adopting a three-dimensional engine for robot training; the virtual laser depth camera of the robot is used as sensor input in the system, the virtual laser depth sensor is installed on the virtual robot, the robot SLAM technology is operated in the ROS (robot operating system) system for autonomous navigation and collection of space and equipment data of the robot, a data set for convolutional neural network training is automatically generated, and a pre-training network is generated through GPU clusters; finally, design survey data is output over the pre-training network, including the number of devices, the type of devices, and the location coordinates of the devices relative to the three-dimensional reconstruction.
As a preferred technical measure:
the SLAM technical target is that under the condition that no priori knowledge exists, a surrounding environment map is built in real time according to virtual laser depth sensor data, and meanwhile, the self positioning is presumed according to the map, and a virtual robot carries a virtual laser depth sensor to move in an unknown environment;
for convenience, changing the motion of a continuous time into discrete moments t=1, … k, wherein at the moments, the position of the virtual robot is denoted by x, and the positions at the moments are denoted by x1, x2 … xk, which form the track of the virtual robot;
in the aspect of the map, the map consists of a plurality of three-dimensional point coordinates, and at each moment, a virtual laser depth sensor measures a part of the three-dimensional point coordinates to obtain observation data of the three-dimensional point coordinates;
the three-dimensional point coordinates are N in number and are represented by y1 and y2 … yn; the localization problem x and mapping problem y are solved by motion measurement u and virtual laser depth sensor reading z.
As a preferred technical measure:
calling a laser ToF sensor, completing three-dimensional point cloud collection of a power distribution room by workers on the scene of the power distribution room, and judging the collected laser point cloud by a simple quantity;
the point cloud with the number of points exceeding 50000 can be used as the input of a room key point coordinate extraction convolutional neural network, and no more than 50000 points require a worker to collect the three-dimensional point cloud again;
after the point cloud is input into the convolutional neural network, the key point coordinates of the basic shape of the room are inferred (information) and vectorized.
The three-dimensional data is collected on site, so that the method can be combined with synchronous positioning and mapping technology, can be independently applied, further can effectively improve data collection efficiency, and meanwhile improves the diversity and accuracy of input data of the convolutional neural network.
As a preferred technical measure:
the field staff shoots a depth image of the existing high-voltage power distribution equipment of the user by adopting the same laser TOF sensor, and the obtained laser depth image firstly passes through an invalid pixel point judging process;
if the number of invalid pixels exceeds 15% of the total number of pixels, a worker is required to retake the high-voltage power distribution equipment. And (3) inputting the depth image with invalid pixel points less than 15% of the total pixel quantity into the pre-training network distribution equipment to identify the depth convolutional neural network after processing the depth image by a bilateral filtering denoising algorithm.
As a preferred technical measure:
completing the steps of equipment classification and position coordinate extraction in the network, if an unrecognizable model exists in output, the on-site staff needs to shoot the depth image again, and restarting the process; if all the models are successfully identified, the quantity, the type and the central point coordinate parameters of the distribution equipment are output.
As a preferred technical measure: the convolutional neural network is a feedforward convolutional neural network, and the CNNC structure of the convolutional neural network is divided into five layers:
the first layer inputs pictures, and Convolution (Convolition) operation is carried out to obtain a matrix (Feature Map) with the depth of the second layer being 3;
pooling (Pooling) is carried out on the matrix of the second layer to obtain a matrix with the depth of 3 of the third layer;
repeating the above operations to obtain a matrix with a fifth layer depth of 5, finally expanding and connecting the 5 matrices according to rows to form vectors, and transmitting the vectors into a full Connected (Connected) layer, wherein the full Connected layer is a BP convolutional neural network; each matrix can be seen as neurons arranged in a matrix form, which are of the same size as neurons in a BP convolutional neural network.
As a preferred technical measure:
setting the number of layers for zero padding according to the requirement; the Zero Padding layer is called Zero Padding, which is a settable super parameter, but the size of the input matrix is adjusted according to the size and the stride of the convolution kernel so that the convolution kernel just slides to the edge;
in general, the input image matrix and the following convolution kernel are square matrices, where the input matrix is w, the convolution kernel is k, the stride is s, and the zero padding layer number is p, and the calculation formula of the feature image size generated after convolution is:
as a preferred technical measure:
in order to extract more features, a plurality of convolution kernels are adopted to respectively carry out convolution, so that a plurality of feature graphs can be obtained; sometimes, for a three-channel color picture, a set of matrices is input, and the convolution kernel is no longer one layer, but rather becomes the corresponding depth.
Compared with the prior art, the invention has the following beneficial effects:
the three-dimensional information and pictures of the power distribution room are collected through the laser depth sensor data, parameters required by drawing are obtained through a convolutional neural network, and the parameters are transmitted to a CAD module; finally outputting a practically available distribution room design drawing; the industrial expansion design time can be effectively shortened, and the industrial expansion design efficiency is improved.
By applying the scheme of the invention, the existing high-voltage industrial expansion design flow can be reduced from the original 5 working days to 0.5 working days, and the revolutionary improvement of the working efficiency is realized. The invention can shorten the design period, standardize engineering standard, increase supply and expense, promote user satisfaction and finally continuously promote the generation of enterprise value by organically integrating power supply service, power engineering design, power engineering construction, power equipment manufacture and the like, thereby forming a business mode of virtuous circle development.
Drawings
FIG. 1 is a flow chart of the operation of the present invention;
FIG. 2 is a flow chart of the operation of the present invention to output device number, device type and location coordinates.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
On the contrary, the invention is intended to cover any alternatives, modifications, equivalents, and variations as may be included within the spirit and scope of the invention as defined by the appended claims. Further, in the following detailed description of the present invention, certain specific details are set forth in order to provide a better understanding of the present invention. The present invention will be fully understood by those skilled in the art without the details described herein.
As shown in fig. 1-2, a cloud computing-based user power distribution room design generation method,
which comprises the following steps:
firstly, establishing an interface program WSGI of a Web server, a Web application program and a depth computing framework which are defined by using a voice for python programming;
secondly, collecting three-dimensional information of the power distribution room by using a laser depth camera through staff on the scene of the power distribution room;
thirdly, transmitting the three-dimensional reconstruction information and the laser depth photo of the equipment to a WSGI of a cloud server through a wireless network;
transmitting data to a convolutional neural network reasoning reference module based on the GPU cluster through a WSGI interface of a cloud, acquiring parameters required by drawing, and transmitting the parameters to a pythonAPI interface of a CAD module;
and fifthly, using a cloud acceleration computing architecture to accelerate the space computation amount of data collected by the site sensor, and finally outputting a standard drawing of the site fund collection of the power distribution room.
The three-dimensional information and pictures of the power distribution room are collected through the laser depth sensor data, parameters required by drawing are obtained through a convolutional neural network, and the parameters are transmitted to a CAD module; finally outputting a practically available distribution room design drawing; the industrial expansion design time can be effectively shortened, and the industrial expansion design efficiency is improved.
By applying the scheme of the invention, the existing high-voltage industrial expansion design flow can be reduced from the original 5 working days to 0.5 working days, and the revolutionary improvement of the working efficiency is realized. The invention can shorten the design period, standardize engineering standard, increase supply and expense, promote user satisfaction and finally continuously promote the generation of enterprise value by organically integrating power supply service, power engineering design, power engineering construction, power equipment manufacture and the like, thereby forming a business mode of virtuous circle development.
The convolutional neural network input source of the invention is a specific embodiment:
the input data of the convolutional neural network is a three-dimensional reconstruction point cloud of a power distribution room, which is collected by a laser depth camera through a synchronous positioning and mapping technology SLAM, a partial differential algorithm is applied to the point cloud, points with large variation are obtained to serve as characteristic point clouds, and then a plurality of coordinate points are randomly extracted to serve as an input source of the convolutional neural network; the device image taken by the laser depth camera is used as another input to the convolutional neural network.
The convolutional neural network training simulation of the invention is a specific embodiment:
training the convolutional neural network, completing data collection through a simulation system, and randomly generating a plurality of power distribution rooms and equipment with different sizes by adopting a three-dimensional engine for robot training; the virtual laser depth camera of the robot is used as sensor input in the system, the virtual laser depth sensor is installed on the virtual robot, the robot SLAM technology is operated in the ROS (robot operating system) system for autonomous navigation and collection of space and equipment data of the robot, a data set for convolutional neural network training is automatically generated, and a pre-training network is generated through GPU clusters; finally, design survey data is output over the pre-training network, including the number of devices, the type of devices, and the location coordinates of the devices relative to the three-dimensional reconstruction.
One embodiment of the SLAM technology of the present invention:
the SLAM technical target is that under the condition that no priori knowledge exists, a surrounding environment map is built in real time according to virtual laser depth sensor data, and meanwhile, the self positioning is presumed according to the map, and a virtual robot carries a virtual laser depth sensor to move in an unknown environment;
for convenience, changing the motion of a continuous time into discrete moments t=1, … k, wherein at the moments, the position of the virtual robot is denoted by x, and the positions at the moments are denoted by x1, x2 … xk, which form the track of the virtual robot;
in the aspect of the map, the map consists of a plurality of three-dimensional point coordinates, and at each moment, a virtual laser depth sensor measures a part of the three-dimensional point coordinates to obtain observation data of the three-dimensional point coordinates;
the three-dimensional point coordinates are N in number and are represented by y1 and y2 … yn; the localization problem x and mapping problem y are solved by motion measurement u and virtual laser depth sensor reading z.
Another specific embodiment of the convolutional neural network input source of the present invention:
calling a laser ToF sensor, completing three-dimensional point cloud collection of a power distribution room by workers on the scene of the power distribution room, and judging the collected laser point cloud by a simple quantity;
the point cloud with the number of points exceeding 50000 can be used as the input of a room key point coordinate extraction convolutional neural network, and no more than 50000 points require a worker to collect the three-dimensional point cloud again;
after the point cloud is input into the convolutional neural network, the key point coordinates of the basic shape of the room are inferred (information) and vectorized.
A specific embodiment of invalid pixel point judgment in the invention is as follows:
the field staff shoots a depth image of the existing high-voltage power distribution equipment of the user by adopting the same laser TOF sensor, and the obtained laser depth image firstly passes through an invalid pixel point judging process;
if the number of invalid pixels exceeds 15% of the total number of pixels, a worker is required to retake the high-voltage power distribution equipment. And (3) inputting the depth image with invalid pixel points less than 15% of the total pixel quantity into the pre-training network distribution equipment to identify the depth convolutional neural network after processing the depth image by a bilateral filtering denoising algorithm.
Completing the steps of equipment classification and position coordinate extraction in the network, if an unrecognizable model exists in output, the on-site staff needs to shoot the depth image again, and restarting the process; if all the models are successfully identified, the quantity, the type and the central point coordinate parameters of the distribution equipment are output.
One specific embodiment of the convolutional neural network architecture hierarchy of the present invention: the convolutional neural network is a feedforward convolutional neural network, and the CNNC structure of the convolutional neural network is divided into five layers:
the first layer inputs pictures, and Convolution (Convolition) operation is carried out to obtain a matrix (Feature Map) with the depth of the second layer being 3;
pooling (Pooling) is carried out on the matrix of the second layer to obtain a matrix with the depth of 3 of the third layer;
repeating the above operations to obtain a matrix with a fifth layer depth of 5, finally expanding and connecting the 5 matrices according to rows to form vectors, and transmitting the vectors into a full Connected (Connected) layer, wherein the full Connected layer is a BP convolutional neural network; each matrix can be seen as neurons arranged in a matrix form, which are of the same size as neurons in a BP convolutional neural network.
Setting the number of layers for zero padding according to the requirement; the Zero Padding layer is called Zero Padding, which is a settable super parameter, but the size of the input matrix is adjusted according to the size and the stride of the convolution kernel so that the convolution kernel just slides to the edge;
in general, the input image matrix and the following convolution kernel are square matrices, where the input matrix is w, the convolution kernel is k, the stride is s, and the zero padding layer number is p, and the calculation formula of the feature image size generated after convolution is:
in order to extract more features, a plurality of convolution kernels are adopted to respectively carry out convolution, so that a plurality of feature graphs can be obtained; sometimes, for a three-channel color picture, a set of matrices is input, and the convolution kernel is no longer one layer, but rather becomes the corresponding depth.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (1)
1. A user power distribution room design generation method based on cloud computing is characterized in that,
which comprises the following steps:
firstly, establishing an interface program WSGI of a Web server, a Web application program and a depth computing framework which are defined by using a voice for python programming;
secondly, collecting three-dimensional information of the power distribution room by using a laser depth camera through staff on the scene of the power distribution room;
thirdly, transmitting the three-dimensional reconstruction information and the laser depth photo of the equipment to a WSGI of a cloud server through a wireless network;
transmitting data to a convolutional neural network reasoning reference module based on the GPU cluster through a WSGI interface of a cloud, acquiring parameters required by drawing, and transmitting the parameters to a pythonAPI interface of a CAD module;
fifthly, using a cloud acceleration computing architecture to accelerate the space computation amount of data collected by the site sensor, and finally outputting a standard drawing of the site collection of the power distribution room;
the input data of the convolutional neural network is a three-dimensional reconstruction point cloud of a power distribution room, which is collected by a laser depth camera through a synchronous positioning and mapping technology SLAM, a partial differential algorithm is applied to the point cloud, points with large variation are obtained to serve as characteristic point clouds, and then a plurality of coordinate points are randomly extracted to serve as an input source of the convolutional neural network; taking an equipment image shot by a laser depth camera as another input of the convolutional neural network;
training the convolutional neural network, completing data collection through a simulation system, and randomly generating a plurality of power distribution rooms and equipment with different sizes by adopting a three-dimensional engine for robot training; the virtual laser depth camera of the robot is used as sensor input in the system, the virtual laser depth sensor is installed on the virtual robot, a robot SLAM technology is operated in the ROS system for autonomous navigation and collection of space and equipment data of the robot, a data set for convolutional neural network training is automatically generated, and a pre-training network is generated through a GPU cluster; finally, outputting design investigation data comprising the number of devices, the types of the devices and the position coordinates of the devices relative to the three-dimensional reconstruction through a pre-training network;
the SLAM technical target is that under the condition that no priori knowledge exists, a surrounding environment map is built in real time according to virtual laser depth sensor data, and meanwhile, the self positioning is presumed according to the map, and a virtual robot carries a virtual laser depth sensor to move in an unknown environment;
changing the motion of a continuous time into discrete moments t=1, … k, wherein at the moments, the self position of the virtual robot is represented by x, and the positions at the moments are marked as x1, x2 … xk, which form the track of the virtual robot;
in the aspect of the map, the map consists of a plurality of three-dimensional point coordinates, and at each moment, a virtual laser depth sensor measures a part of the three-dimensional point coordinates to obtain observation data of the three-dimensional point coordinates;
the three-dimensional point coordinates are N in number and are represented by y1 and y2 … yn; solving a positioning problem x and a mapping problem y through motion measurement u and virtual laser depth sensor reading z;
calling a laser ToF sensor, completing three-dimensional point cloud collection of a power distribution room by workers on the scene of the power distribution room, and judging the collected laser point cloud by a simple quantity;
the point cloud with the number of points exceeding 50000 can be used as the input of a room key point coordinate extraction convolutional neural network, and no more than 50000 points require a worker to collect the three-dimensional point cloud again;
after the point cloud is input into the convolutional neural network, deducing the key point coordinates of the basic shape of the room and vectorizing;
the field staff shoots a depth image of the existing high-voltage power distribution equipment of the user by adopting the same laser TOF sensor, and the obtained laser depth image firstly passes through an invalid pixel point judging process;
if the number of the invalid pixels exceeds 15% of the total number of the pixels, a worker is required to shoot the high-voltage power distribution equipment again; the depth image with invalid pixel points less than 15% of the total pixel is processed by a bilateral filtering denoising algorithm and then input into a pre-training distribution network device to identify a depth convolutional neural network;
completing the steps of equipment classification and position coordinate extraction in the network, if an unrecognizable model exists in output, the on-site staff needs to shoot the depth image again, and restarting the process; if all the models are successfully identified, outputting the quantity, the type and the central point coordinate parameters of the power distribution equipment; the convolutional neural network is a feedforward convolutional neural network, and the CNNC structure of the convolutional neural network is divided into five layers:
inputting pictures in the first layer, and performing convolution operation to obtain a matrix with the depth of the second layer being 3;
pooling the matrix of the second layer to obtain a matrix with the depth of 3 of the third layer;
repeating the above operations to obtain a matrix with a fifth layer depth of 5, and finally expanding and connecting the 5 matrices according to rows to form vectors, and transmitting the vectors into a full-connection layer, wherein the full-connection layer is a BP convolutional neural network; each matrix can be regarded as neurons arranged in a matrix form, and the sizes of the neurons in the matrix form are different from those of the neurons in the BP convolutional neural network;
setting the number of layers for zero padding according to the requirement; the Zero Padding layer is called Zero Padding, which is a settable super parameter, but the size of the input matrix is adjusted according to the size and the stride of the convolution kernel so that the convolution kernel just slides to the edge;
the input image matrix and the following convolution kernel are square matrixes, wherein the size of the input matrix is w, the size of the convolution kernel is k, the stride is s, the zero padding layer number is p, and the calculation formula of the size of the feature image generated after convolution is as follows:
；
in order to extract more features, a plurality of convolution kernels are adopted to respectively carry out convolution, so that a plurality of feature graphs can be obtained; for a three-channel color picture, a set of matrices is input, and the convolution kernel is no longer one layer but is changed to a corresponding depth.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911341997.6A CN111192363B (en) | 2019-12-23 | 2019-12-23 | User power distribution room design generation method based on cloud computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911341997.6A CN111192363B (en) | 2019-12-23 | 2019-12-23 | User power distribution room design generation method based on cloud computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111192363A CN111192363A (en) | 2020-05-22 |
| CN111192363B true CN111192363B (en) | 2023-09-29 |
Family
ID=70707457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911341997.6A Active CN111192363B (en) | 2019-12-23 | 2019-12-23 | User power distribution room design generation method based on cloud computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111192363B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111832542B (en) * | 2020-08-15 | 2024-04-16 | 武汉易思达科技有限公司 | Tri-vision identifying and positioning device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400415A (en) * | 2013-08-07 | 2013-11-20 | 赞奇科技发展有限公司 | Implementation method and system for automatic generation and on-line interaction of three-dimensional digital model with planar space structure |
| CN105027190A (en) * | 2013-01-03 | 2015-11-04 | 美达公司 | Extramissive spatial imaging digital eye glass for virtual or augmediated vision |
| CN108171796A (en) * | 2017-12-25 | 2018-06-15 | 燕山大学 | A kind of inspection machine human visual system and control method based on three-dimensional point cloud |
| CN108365994A (en) * | 2018-03-13 | 2018-08-03 | 山东超越数控电子股份有限公司 | A kind of cloud security management platform for cloud computing security and unity management |
| CN108921893A (en) * | 2018-04-24 | 2018-11-30 | 华南理工大学 | A kind of image cloud computing method and system based on online deep learning SLAM |
| CN109801358A (en) * | 2018-12-06 | 2019-05-24 | 宁波市电力设计院有限公司 | A kind of substation's three-dimensional investigation method scanning and put cloud visual fusion based on SLAM |
-
2019
- 2019-12-23 CN CN201911341997.6A patent/CN111192363B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105027190A (en) * | 2013-01-03 | 2015-11-04 | 美达公司 | Extramissive spatial imaging digital eye glass for virtual or augmediated vision |
| CN103400415A (en) * | 2013-08-07 | 2013-11-20 | 赞奇科技发展有限公司 | Implementation method and system for automatic generation and on-line interaction of three-dimensional digital model with planar space structure |
| CN108171796A (en) * | 2017-12-25 | 2018-06-15 | 燕山大学 | A kind of inspection machine human visual system and control method based on three-dimensional point cloud |
| CN108365994A (en) * | 2018-03-13 | 2018-08-03 | 山东超越数控电子股份有限公司 | A kind of cloud security management platform for cloud computing security and unity management |
| CN108921893A (en) * | 2018-04-24 | 2018-11-30 | 华南理工大学 | A kind of image cloud computing method and system based on online deep learning SLAM |
| CN109801358A (en) * | 2018-12-06 | 2019-05-24 | 宁波市电力设计院有限公司 | A kind of substation's three-dimensional investigation method scanning and put cloud visual fusion based on SLAM |
Non-Patent Citations (4)
| Title |
|---|
| 张向荣等.《模式识别》.西安电子科技大学出版社,2019,第161-164页. * |
| 张彬等.《图像复原优化算法》.国防工业出版社,2019,第262页. * |
| 王权等.基于ROS与Kinect的移动机器人同时定位与地图构建.《计算机应用研究》.2016,第34卷(第10期),第3184-3187页. * |
| 黑马程序员.《Python程序开发案例教程》.中国铁道出版社有限公司,2019,第262页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111192363A (en) | 2020-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110458939B (en) | Indoor scene modeling method based on visual angle generation | |
| WO2021175050A1 (en) | Three-dimensional reconstruction method and three-dimensional reconstruction device | |
| CN111598998A (en) | Three-dimensional virtual model reconstruction method and device, computer equipment and storage medium | |
| CN108154104A (en) | A kind of estimation method of human posture based on depth image super-pixel union feature | |
| CN110276768B (en) | Image segmentation method, image segmentation device, image segmentation apparatus, and medium | |
| CN104463969B (en) | A kind of method for building up of the model of geographical photo to aviation tilt | |
| CN111104702A (en) | Method for realizing visual simulation of embankment project based on UE4 | |
| CN113822982A (en) | Human body three-dimensional model construction method and device, electronic equipment and storage medium | |
| CN111028335B (en) | Point cloud data block surface patch reconstruction method based on deep learning | |
| CN111192363B (en) | User power distribution room design generation method based on cloud computing | |
| CN110176063B (en) | Clothing deformation method based on human body Laplace deformation | |
| Petkov et al. | Interactive visibility retargeting in vr using conformal visualization | |
| CN109064482B (en) | Method and device for automatically acquiring house outline in three-dimensional oblique photography scene | |
| CN116797733A (en) | Real-time three-dimensional object dynamic reconstruction method | |
| CN111047703B (en) | User high-voltage distribution equipment identification and space reconstruction method | |
| CN111368637A (en) | Multi-mask convolution neural network-based object recognition method for transfer robot | |
| CN113284249B (en) | Multi-view three-dimensional human body reconstruction method and system based on graph neural network | |
| CN112002019B (en) | Method for simulating character shadow based on MR mixed reality | |
| CN114419255A (en) | Three-dimensional human head model generation method and device fusing real human faces, electronic equipment and storage medium | |
| CN110689616B (en) | Water delivery channel parametric modeling method based on three-dimensional digital earth | |
| Sun et al. | Weakly-supervised reconstruction of 3D objects with large shape variation from single in-the-wild images | |
| CN112837420A (en) | Method and system for completing shape of terracotta warriors point cloud based on multi-scale and folding structure | |
| CN113538686B (en) | Regional building 3D modeling method, device, equipment and storage medium | |
| CN116958478B (en) | City building model programming generation method, device, equipment and storage medium | |
| CN116912433B (en) | Three-dimensional model skeleton binding method, device, 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 |