CN111611937A - Prison personnel abnormal behavior monitoring method based on BIM and neural network - Google Patents
Prison personnel abnormal behavior monitoring method based on BIM and neural network Download PDFInfo
- Publication number
- CN111611937A CN111611937A CN202010442776.4A CN202010442776A CN111611937A CN 111611937 A CN111611937 A CN 111611937A CN 202010442776 A CN202010442776 A CN 202010442776A CN 111611937 A CN111611937 A CN 111611937A
- Authority
- CN
- China
- Prior art keywords
- prison
- bim
- neural network
- personnel
- images
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 26
- 238000012544 monitoring process Methods 0.000 title claims abstract description 13
- 206010000117 Abnormal behaviour Diseases 0.000 title claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 17
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 13
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 230000000007 visual effect Effects 0.000 claims abstract description 6
- 238000010586 diagram Methods 0.000 claims description 37
- 230000006399 behavior Effects 0.000 claims description 13
- 230000009466 transformation Effects 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000001931 thermography Methods 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Tourism & Hospitality (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Alarm Systems (AREA)
Abstract
The invention provides a prison personnel abnormal behavior monitoring method based on BIM and a neural network, which comprises the following steps: building BIM of the prison, building a virtual prison three-dimensional model, and providing a building engineering information base for the model; establishing a deep neural network, wherein the deep neural network comprises an image acquisition processing unit, a heat point regression unit and a behavior analysis unit; the block chain and the encryption algorithm are adopted, the neural network is divided into blocks, data transmitted among the blocks are encrypted, and the data are prevented from being stolen and tampered by others; and transmitting the result obtained by the neural network to a BIM of the prison for storage, performing visual processing through a Web GIS, and displaying the BIM of the prison and the distribution state of personnel in the prison on the Web, so that supervisors can know the abnormal walking condition of on-duty dry police or prisoners in a prison area and take corresponding management measures. The method can enable the supervision personnel to better carry out supervision work.
Description
Technical Field
The invention relates to the field of artificial intelligence and block chains, in particular to a prison personnel abnormal behavior monitoring method based on BIM and a neural network.
Background
With the continuous development of scientific technology, for prison staff, the traditional supervision and management method based on human resources, namely 'people stare at and people see people' has many limitations, which not only consumes a lot of human resources, but also leaves the question about the effectiveness of supervision because the supervision is affected by many factors such as attention and visual field dead zones.
In the current proposed intelligent prison construction scheme, part of facilities still rely on sensors to obtain position information data, and then the data is sent to a server for integration and processing. The limitations of the sensors are: a position sensor must be equipped for each prisoner and prison police, and the sensors need to be supplied with power in time; due to the particularity of the prison system, the internal network of the prison system does not have the condition of high-speed optical fiber for information transmission, and the information is processed and only stored by a memory, so that the prison system has great potential safety hazard.
Disclosure of Invention
In order to solve the problems of the existing monitoring method, the invention provides a prison personnel abnormal behavior monitoring method based on BIM and a neural network, which comprises the following steps:
building BIM of the prison, building a virtual prison three-dimensional model, and providing a building engineering information base for the model;
establishing a deep neural network, wherein the deep neural network comprises an image acquisition processing unit, a heat point regression unit and a behavior analysis unit;
acquiring images of personnel distribution conditions in the prison through a camera included in an image acquisition and processing unit, and preprocessing the acquired images in the unit;
the preprocessed image passes through an encoder and a decoder in a thermal point regression unit to obtain a thermodynamic diagram;
sending the thermodynamic diagram into a behavior analysis unit, wherein the behavior analysis unit comprises two branches, building geometric information of prisons in BIM is combined, and a first branch outputs a detection result of the boundary crossing condition of people; the second branch is used for judging whether the personnel have abnormal staying behaviors or not, and outputting coordinates of the abnormal staying positions if the personnel have the abnormal staying behaviors;
and transmitting the results obtained by the two branches to the BIM of the prison for storage, performing visual processing through the WebGIS, and displaying the BIM of the prison and the personnel distribution state in the prison on Web.
The information base comprises the shape, size, spatial distribution and personnel distribution information of the prisons.
The camera comprises an RGB camera and a thermal imaging camera with similar poses.
And preprocessing, namely projecting the images acquired by the two cameras into the same plane through projection transformation, performing image registration operation on the images, and aligning the images to obtain images with the same size.
The encoder is used for performing convolution on the preprocessed image to extract features, and the decoder is used for decoding the output of the encoder to obtain the thermal diagram.
The first branch is used for performing softargmax operation on the thermodynamic diagram to obtain coordinates of peak points in each hotspot, projecting the coordinates into a virtual three-dimensional model of the prison, and outputting a detection result of the personnel border crossing condition by combining information in a BIM information base; the second branch comprises a superposition module; sending the thermodynamic diagrams into a superposition unit for storage according to a time sequence of one second frame, after a set frame number is reached, carrying out superposition operation based on a forgetting algorithm on the thermodynamic diagrams, setting a threshold value, judging abnormal stay when the peak point of the new thermodynamic diagrams obtained after superposition is higher than the threshold value, carrying out softargmax processing on the thermodynamic diagrams obtained after superposition to obtain the coordinates of the peak point of the thermodynamic, projecting the coordinates into a virtual three-dimensional model of the prison, and combining information included in a BIM information base to obtain the coordinates of the abnormal stay position.
The forgetting algorithm has a formula of X ═ α X + (1- α) X ', where X is a current frame thermodynamic diagram, X' is a result obtained by superimposing thermodynamic diagrams before the current frame, X is a result obtained by superimposing the thermodynamic diagrams, and 1- α is a forgetting coefficient.
The images collected by the two cameras are used as a training data set, a point projected to the ground by a person is subjected to Gaussian fuzzy operation, hot spot labeling is carried out on the point, and a depth neural network is trained by using a mean square error loss function.
The thermal point regression unit and the behavior analysis unit based on the deep neural network are mainly composed of subtask blocks, each subtask block comprises an encoder, a decoder and a superposition module, each subtask block corresponds to a block, nodes are randomly distributed to the blocks according to a random number sequence, the image acquisition processing unit is a block, a block chain private chain is generated according to the inference sequence of the network, and data transmitted among the blocks are encrypted.
And encrypting the data transmitted among the blocks by using an encryption algorithm, decrypting the encrypted data by using a corresponding decryption algorithm before the encrypted data flows into the next block, and executing the encryption and decryption processes until the monitoring is finished.
The invention has the beneficial effects that:
1. by adopting the method, the waste of human resources can be effectively reduced, the method is not influenced by subjective factors of supervisors, abnormal conditions can be found in time, irrecoverable loss is avoided, the limitation of the traditional supervision method mainly relying on human resources is broken through, and the effectiveness of the prison supervision system is improved.
2. The method adopts the deep neural network technology, improves the inconvenience that the sensor needs to be charged in time when used in the prior art, and leads the supervision work of related personnel to be more efficient and convenient; the block chain and the encryption technology provide guarantee for the safety of related information data, and can effectively prevent someone with no intention from tampering and stealing the data.
3. The method is based on BIM and Web GIS technology, can quickly and intuitively know the occurrence position of the abnormal situation, and is convenient for supervisors to know the abnormal walking situation of on-duty police officers or prisoners in prison areas and take corresponding management measures.
Drawings
Fig. 1 is a block diagram of a deep neural network.
Detailed Description
The method for monitoring the abnormal behavior of the prison personnel based on the BIM and the neural network is further described in the following by combining the embodiment and the attached drawings, and the method is shown in figure 1.
Example (b):
a prison personnel abnormal behavior monitoring method based on BIM and neural network comprises the following steps:
building BIM of the prison: according to prisons in real life, real information of the prisons is simulated through digital information simulation, and a virtual three-dimensional model of the prisons is constructed.
The core of BIM is to provide a complete building engineering information base consistent with the actual situation for a virtual building engineering three-dimensional model by establishing the model and utilizing the digitization technology, wherein the information base comprises the shape, the size, the spatial distribution and the personnel distribution information of prisons.
The model not only comprises the geometrical components of the prison, such as beams, columns, walls, doors, windows and the like, but also comprises the spatial relationship among the geometrical components, the number of the components and attribute data, such as color, material and the like.
Monitoring the abnormal behaviors of prisoners by adopting a deep neural network and a block chain technology, wherein the frame of the deep neural network is shown in figure 1 and comprises the following steps:
the image acquisition and processing unit acquires images of personnel positions in corresponding areas through the RGB cameras and the thermal imaging cameras which are close in position and orientation in each area of the prison, preprocesses the acquired images, namely projects the images acquired by the two cameras into the same plane through projection transformation, performs image registration operation on the images, aligns the images to obtain the images with the same size, and combines the two preprocessed images to serve as the output of the unit.
Images shot by two cameras with similar poses are not in the same plane, and in order to enable a network to better analyze and operate the images, the acquired images need to be projected into the same plane through projection transformation; projection transformation, also called homography transformation, is a transformation of coordinates of a map projection point into coordinates of another map projection point, and plays a very important role in the fields of image correction, image stitching, camera pose estimation and the like.
Image registration, which is a process of matching and superimposing two or more images acquired at different times and different sensors (imaging devices) or under different conditions (weather, illuminance, camera position, angle, and the like); specifically, for two images in a set of image data sets, one image is mapped to the other image by finding a spatial transformation, so that points corresponding to the same position in space in the two images are in one-to-one correspondence, and the purpose of information fusion is achieved.
The thermal point regression unit comprises an encoder and a decoder, wherein the encoder performs convolution on the image output by the image acquisition processing unit to extract features, and the decoder decodes the output of the encoder to obtain a thermodynamic diagram.
A behavior analysis unit comprising two branches, wherein:
the first branch is used for performing softargmax operation on the thermodynamic diagram output by the thermal point regression unit to obtain the coordinates of each peak point. Projecting the coordinates into a virtual three-dimensional model of the prison by utilizing a camera matrix, and outputting a detection result of the personnel border crossing condition by combining the prison building information contained in the BIM;
the second branch comprises a superposition module, the thermodynamic diagrams output by the thermodynamic point regression unit are sent to the superposition module for storage according to a time sequence of one second and one frame, after the set number of frames is reached, superposition operation based on a forgetting algorithm is carried out on the thermodynamic diagrams, a threshold value is set, abnormal stay is judged when the value of the thermodynamic diagrams after superposition is larger than the threshold value, softargmax processing is carried out on the thermodynamic diagrams after superposition, coordinates are projected into a virtual three-dimensional model of prisons by using a camera matrix, and coordinates of abnormal stay positions are obtained by combining prison building information included in BIM.
Forgetting algorithm: the calculation formula is X ═ α X + (1- α) X ', where X is the thermodynamic diagram of the current frame, X' is the result after the thermodynamic diagrams before the current frame are superimposed, X is the result of superimposing the thermodynamic diagrams, 1- α is the forgetting coefficient, and α in the embodiment is 0.05.
The camera matrix is also called a camera projection matrix and can be decomposed into a product of two matrixes of an internal reference matrix and an external reference matrix, and the product is used for establishing a projection relation between three dimensions and two dimensions.
Training of the network:
collecting a plurality of frames of RGB and thermal imaging images in each area of the prison for a period of time as a training data set; performing Gaussian fuzzy operation on a point projected to the ground by a person in the image, and performing hot spot labeling on the point; a mean square error loss function is taken.
Gaussian blurring, also known as gaussian smoothing, is commonly used to reduce image noise and reduce levels of detail; the operation process is to convolute the image with the normal distribution.
The Mean Square Error (MSE), which is the most commonly used error in the regression loss function, is the sum of the squares of the difference between the predicted value and the target value, and is given by the following formula:
wherein, YiIs the target value, f (X)i) An estimate is output for the model.
In the prior art, the processing of information is completed on a computing cluster, data is easy to leak, and great potential safety hazard is caused, so that the method adopts a block chain and an encryption technology to carry out module splitting on a network, and carries out encryption processing on data transmitted among blocks, thereby ensuring the safety of related data and avoiding irreparable loss.
The image acquisition and processing unit is a block, generates a block chain private chain according to the inference sequence of the network, and encrypts data transmitted among the blocks.
The image acquisition processing unit is a block 1, the encoder is a block 2, the decoder is a block 3, and the superposition module is a block 4; the reasoning sequence is as follows: the block 1 is used for collecting images and preprocessing the images, the preprocessed images are sent to a block 2 to be convoluted, the obtained characteristics are used as input of the block 3, the block 3 decodes the input to obtain thermodynamic diagrams, the obtained thermodynamic diagrams are sent to a block 4 for storing one frame in one second according to time sequence, when a certain number of frames is reached, the stored thermodynamic diagrams are superposed, when the heat peak point of a heat spot in the thermodynamic diagrams is higher than a preset threshold value, the heat peak point is stopped for a person abnormally, and the superposed thermodynamic diagrams are processed to obtain heat peak point coordinates.
A random number generation mechanism: firstly, a new random number seed is generated, and after the random number seed is selected, the random number is generated by using a linear feedback shift register. Setting seed as X0The method comprises calculating the seed with a linear function to obtain input, performing XOR operation on some bits to obtain new output, shifting the whole register, and repeating the steps to obtain a series of random numbers in the range of [1, N ]]And N is the number of nodes. The random number seed can be generated by various methods such as IP, Global Time, etc., and can be selected by the implementer.
Randomly selecting nodes in a local computing cluster: and sequencing available nodes in the cluster, wherein each node has a fixed serial number, distributing a series of generated random numbers to the blocks 2-4, and selecting the nodes with the same serial numbers as the random numbers distributed to the blocks for operation by the blocks.
The embodiment encrypts the data transmitted among the blocks by adopting an AES symmetric encryption algorithm, thereby ensuring the running speed of data encryption transmission. The implementer can choose the encryption algorithm by himself.
The plaintext and the key pass through an AES encryption function to obtain a ciphertext, the ciphertext and the key pass through the AES decryption function to obtain the plaintext before the encrypted data flow into the next block, and the above processes are executed until the monitoring is completed; specifically, the plaintext is divided into a group, the length of the group is 128 bits, and one group of data is encrypted each time until the whole plaintext is encrypted; the length of the key can use 128 bits, 192 bits or 256 bits, the length of the key is different, and the number of encryption rounds is also different; a round function is performed in the encryption function and the decryption function, which includes four operations: byte substitution, row shift, column mix, round key addition. Before the first round of processing in the encryption process and the first round of processing in the decryption process, the plaintext and the ciphertext need to be subjected to exclusive or operation with the original key respectively, and the column mixing operation is not performed in the last round of processing in the two processes.
Thus, the reasoning of the deep neural network is completed.
And transmitting the operation result of the neural network to the BIM of the prison for storage, and performing visual processing through the WebGIS to realize the display of the BIM of the prison and the states of the personnel in the prison on the Web.
And projecting the target coordinates obtained by two-dimensional image processing in the first branch and the second branch into the BIM of the prison through a matrix, and combining the coordinate information of the camera head, so that the position of the target can be accurately positioned. Meanwhile, visual processing is carried out through the WebGIS, and supervisors can search, inquire and analyze on Web, so that the supervisors can know abnormal walking conditions of on-duty police officers or prisoners in the prison area and take corresponding management measures.
The above-described embodiments are intended to better understand the present invention by those skilled in the relevant art and are not intended to limit the present invention.
Claims (10)
1. A prison personnel abnormal behavior monitoring method based on BIM and neural network is characterized by comprising the following steps:
building BIM of the prison, building a virtual prison three-dimensional model, and providing a building engineering information base for the model;
establishing a deep neural network, wherein the deep neural network comprises an image acquisition processing unit, a heat point regression unit and a behavior analysis unit;
acquiring images of personnel distribution conditions in the prison through a camera included in an image acquisition and processing unit, and preprocessing the acquired images in the unit;
the preprocessed image passes through an encoder and a decoder in a thermal point regression unit to obtain a thermodynamic diagram;
sending the thermodynamic diagram into a behavior analysis unit, wherein the behavior analysis unit comprises two branches, building geometric information of prisons in BIM is combined, and a first branch outputs a detection result of the boundary crossing condition of people; the second branch is used for judging whether the personnel have abnormal staying behaviors or not, and if the personnel have the abnormal staying behaviors, outputting abnormal staying position coordinates;
and transmitting the results obtained by the two branches to the BIM of the prison for storage, performing visual processing through a Web GIS, and displaying the BIM of the prison and the personnel distribution state in the prison on the Web.
2. The method of claim 1, wherein the library of information comprises shape, size, spatial distribution, people distribution information for prisons.
3. The method of claim 2, wherein the cameras comprise both RGB cameras and thermal imaging cameras with similar poses.
4. The method of claim 3, wherein the preprocessing is to project the images collected by the two cameras into the same plane through a projection transformation, and perform an image registration operation on the images to align the images to obtain images with equal size.
5. The method of claim 4, wherein the encoder is configured to convolve the preprocessed images to extract features, and the decoder is configured to decode the output of the encoder to obtain the thermodynamic diagram.
6. The method of claim 5, wherein the first branch is used for performing softargmax operation on the thermodynamic diagram to obtain coordinates of peak points in each hotspot, projecting the coordinates into a virtual three-dimensional model of the prison, and outputting a detection result of the personnel boundary crossing condition by combining information included in a BIM information base; the second branch comprises a superposition module; sending the thermodynamic diagrams into a superposition unit for storage according to a time sequence of one second frame, after the number of the frames reaches a set frame number, carrying out superposition operation based on a forgetting algorithm on the thermodynamic diagrams, setting a threshold value, judging abnormal stay when the peak point of the new thermodynamic diagrams obtained after superposition is higher than the threshold value, carrying out softargmax processing on the thermodynamic diagrams obtained after superposition to obtain the coordinates of the peak point of the thermodynamic, projecting the coordinates into a virtual three-dimensional model of the prison, and combining information in a BIM information base to obtain the coordinates of the abnormal stay position.
7. The method according to claim 6, wherein the forgetting algorithm has a formula of X ═ α X + (1- α) X ', where X is a current frame thermodynamic diagram, X' is a result after superposition of thermodynamic diagrams before the current frame, X is a result of superposition of thermodynamic diagrams, and 1- α is a forgetting coefficient.
8. The method of claim 7, wherein the images collected by the two cameras are used as a training data set, a Gaussian blur operation is performed on a point projected by the person to the ground, hot spot labeling is performed on the point, and a mean square error loss function is used for training the deep neural network.
9. The method according to claim 8, wherein the deep neural network-based thermal point regression unit and the behavior analysis unit are mainly composed of subtask blocks, each subtask block includes an encoder, a decoder, and a superposition module, each subtask block corresponds to a block, nodes are randomly allocated to the blocks according to a random number sequence, the image acquisition processing unit is a block, a private chain of the block chain is generated according to an inference sequence of the network, and data transmitted between the blocks is encrypted.
10. The method of claim 9, wherein the data transmitted between blocks is encrypted using an encryption algorithm, and the encrypted data is decrypted using a corresponding decryption algorithm before flowing into the next block, and the encryption and decryption processes are performed until the monitoring is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010442776.4A CN111611937A (en) | 2020-05-22 | 2020-05-22 | Prison personnel abnormal behavior monitoring method based on BIM and neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010442776.4A CN111611937A (en) | 2020-05-22 | 2020-05-22 | Prison personnel abnormal behavior monitoring method based on BIM and neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111611937A true CN111611937A (en) | 2020-09-01 |
Family
ID=72199590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010442776.4A Withdrawn CN111611937A (en) | 2020-05-22 | 2020-05-22 | Prison personnel abnormal behavior monitoring method based on BIM and neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111611937A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112287873A (en) * | 2020-11-12 | 2021-01-29 | 广东恒电信息科技股份有限公司 | Judicial service early warning system |
| CN112613359A (en) * | 2020-12-09 | 2021-04-06 | 苏州玖合智能科技有限公司 | Method for constructing neural network for detecting abnormal behaviors of people |
| CN112651069A (en) * | 2020-12-05 | 2021-04-13 | 重庆源道建筑规划设计有限公司 | Intelligent construction site management and control method, system and device based on BIM and storage medium |
| CN114727064A (en) * | 2022-04-02 | 2022-07-08 | 清华大学 | Construction safety macro monitoring system and method |
| CN116228467A (en) * | 2023-05-06 | 2023-06-06 | 成都大前研软件开发有限公司 | Power supply method, system, equipment and medium of high-voltage power grid based on artificial intelligence |
-
2020
- 2020-05-22 CN CN202010442776.4A patent/CN111611937A/en not_active Withdrawn
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112287873A (en) * | 2020-11-12 | 2021-01-29 | 广东恒电信息科技股份有限公司 | Judicial service early warning system |
| CN112651069A (en) * | 2020-12-05 | 2021-04-13 | 重庆源道建筑规划设计有限公司 | Intelligent construction site management and control method, system and device based on BIM and storage medium |
| CN112613359A (en) * | 2020-12-09 | 2021-04-06 | 苏州玖合智能科技有限公司 | Method for constructing neural network for detecting abnormal behaviors of people |
| CN112613359B (en) * | 2020-12-09 | 2024-02-02 | 苏州玖合智能科技有限公司 | Construction method of neural network for detecting abnormal behaviors of personnel |
| CN114727064A (en) * | 2022-04-02 | 2022-07-08 | 清华大学 | Construction safety macro monitoring system and method |
| CN116228467A (en) * | 2023-05-06 | 2023-06-06 | 成都大前研软件开发有限公司 | Power supply method, system, equipment and medium of high-voltage power grid based on artificial intelligence |
| CN116228467B (en) * | 2023-05-06 | 2023-11-03 | 国网浙江省电力有限公司丽水供电公司 | Power supply method, system, equipment and medium of high-voltage power grid based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111611937A (en) | Prison personnel abnormal behavior monitoring method based on BIM and neural network | |
| Pan et al. | Cross-view semantic segmentation for sensing surroundings | |
| CN103795976B (en) | A kind of full-time empty 3 d visualization method | |
| CN1332349C (en) | Person tagging in an image processing system utilizing a statistical model based on both appearance and geometric features | |
| Gao et al. | Color-gray multi-image hybrid compression–encryption scheme based on BP neural network and knight tour | |
| CN103260015B (en) | Based on the three-dimensional visible supervisory control system of RGB-Depth camera | |
| CN111814238A (en) | BIM real-time imaging method for breeding house based on artificial intelligence and mixed cloud reasoning | |
| CN111583068A (en) | Statistical method for stay time and heat degree of personnel in construction area combined with block chain | |
| CN110428449A (en) | Target detection tracking method, device, equipment and storage medium | |
| CN105208325A (en) | Territorial resource monitoring and early warning method based on image fixed-point snapshot and comparative analysis | |
| CN109191366A (en) | Multi-angle of view human body image synthetic method and device based on human body attitude | |
| CN111583661A (en) | Vehicle red light running detection method based on gun-ball linkage and DNN | |
| CN111414499A (en) | Operation personnel safety belt wearing detection system based on block chain and BIM | |
| Shiotsuka et al. | Gan-based semantic-aware translation for day-to-night images | |
| CN104539875A (en) | Video monitoring method and system | |
| Chen et al. | Cross-drone transformer network for robust single object tracking | |
| AU2020104112A4 (en) | A RNN based Spatio Temporal Data Mining model for urban road Planning | |
| CN109977856A (en) | Complex behavior knows method for distinguishing in a kind of multisource video | |
| CN111461074A (en) | Channel blockage detection method combining block chain and deep learning | |
| Lu et al. | Bi-temporal attention transformer for building change detection and building damage assessment | |
| CN111523434A (en) | Building site personnel abnormal gathering detection system based on block chain and BIM | |
| CN111476208A (en) | Intelligent building site constructor gathering monitoring system utilizing block chains | |
| CN104517292A (en) | Multi-camera high-density crowd partitioning method based on planar homography matrix restraint | |
| Hu et al. | A spatiotemporal intelligent framework and experimental platform for urban digital twins | |
| CN111414500A (en) | Steel wire rope breakage early warning system based on block chain and BIM |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200901 |
|
| WW01 | Invention patent application withdrawn after publication |