CN110730352A - Power transmission line image decoding method based on variational self-coding - Google Patents
Power transmission line image decoding method based on variational self-coding Download PDFInfo
- Publication number
- CN110730352A CN110730352A CN201911101962.5A CN201911101962A CN110730352A CN 110730352 A CN110730352 A CN 110730352A CN 201911101962 A CN201911101962 A CN 201911101962A CN 110730352 A CN110730352 A CN 110730352A
- Authority
- CN
- China
- Prior art keywords
- transmission line
- power transmission
- server
- coding
- pictures
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/182—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being a pixel
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention relates to the field of power transmission line image generation, in particular to a power transmission line image decoding method based on variational self-coding. A, a camera captures pictures in a scene range of a power transmission line, and the pictures are coded by using a variational self-coding technology; b. and transmitting the data obtained by coding to a server through a 4G network, and reconstructing the codes stored in the server into pictures according to requirements. The invention can realize the quick and low-cost transmission and storage of the monitoring image of the power transmission line, saves the transmission bandwidth and the storage server cost, and is beneficial to realizing the intelligent management of the operation of the power grid.
Description
Technical Field
The invention relates to the field of power transmission line image generation, in particular to a power transmission line image decoding method based on variational self-coding.
Background
In the power industry, the safety problem of the transmission line is a crucial problem. Along with the development of artificial intelligence, the mode of manual inspection is gradually replaced by the intelligent system of the power transmission line. After the cameras capture the pictures of the power transmission line, the captured pictures are directly transmitted to the background server through the 4G network, the number of the pictures captured by each camera every day is 100, a large number of cameras generate a large amount of data every day, and if the pictures are transmitted completely, the network bandwidth is greatly consumed. On the other hand, after the picture data set is uploaded to the server, the data volume is too large, so that a large amount of storage space of the server is occupied, and a large economic cost is also needed.
In summary, a problem to be solved by technical personnel in the field is how to provide an efficient and reliable power transmission line image coding method based on variational self-coding to code and compress an image, and better reconstruct the image through coding, so as to provide technical support for transmitting the image of a power transmission line monitoring image quickly and at low cost.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide the power transmission line image coding and decoding method based on variational self-coding, which can realize the quick low-cost transmission and storage of the power transmission line monitoring image, save the transmission bandwidth and the stored server cost and is beneficial to realizing the intelligent management of the power grid operation.
In order to achieve the above object, the present invention provides a method for decoding transmission line images based on variational self-encoding, comprising the following steps,
a. the method comprises the following steps that a camera captures pictures in a scene range of the power transmission line, and the pictures are coded by using a variational self-coding technology;
b. and transmitting the data obtained by coding to a server through a 4G network, and reconstructing the codes stored in the server into pictures according to requirements.
Preferably, the variational self-coding technology in step a is based on a generation model of an encoder-decoder mode, a neural network is used in the mobile terminal device to encode the picture, and the encoded data is decoded in the server as required to reconstruct a visualized image.
Preferably, step b specifically comprises:
b 1: based on the pictures in the range of the power transmission line in the step a, acquiring hidden danger picture codes by using a variational self-coding technology;
b 2: transmitting the picture codes to a server side through a 4G network;
b 3: collecting the scene hidden danger picture codes of the power transmission line at a server side, and utilizing the obtained scene hidden danger picture codes of the power transmission line;
b 4: and according to the user requirements, reconstructing the codes stored in the server into pictures.
Preferably, b2 is specifically: the camera system comprises an access and control device of a 4G network, the control device automatically uploads a code obtained by decoding a monitoring picture captured by a camera to a server, the access device of the 4G network can be connected to a wireless network of a telecommunication operator to remotely transmit the code, and the generated code is transmitted to a background server through the 4G network.
Preferably, the b4 specifically comprises: the stored power line image is encoded as input to a decoder to generate new samples that are similar but different from the training data.
The invention has the beneficial effects that:
(1) the invention uses the variational self-coding VAE algorithm, and performs targeted improvement work, can realize high-power image compression, can reconstruct an image with better visualization effect, and can effectively reduce the consumption of bandwidth.
(2) The invention effectively saves the storage space of the server by storing the coding information of the image in the server.
Drawings
FIG. 1: the method steps of the invention are flow chart;
FIG. 2: the integral structure diagram of the variational self-coding technology;
FIG. 3: an encoder-decoder skeleton diagram.
Detailed Description
Example 1:
referring to the attached drawings in the specification, the invention provides a power transmission line image decoding method based on variational self-coding, which comprises the following steps,
a. the method comprises the following steps that a camera captures pictures in a scene range of the power transmission line, and the pictures are coded by using a variational self-coding technology;
b. and transmitting the data obtained by coding to a server through a 4G network, and reconstructing the codes stored in the server into pictures according to requirements.
Preferably, the variational self-coding technology in step a is based on a generation model of an encoder-decoder mode, a neural network is used in the mobile terminal device to encode the picture, and the encoded data is decoded in the server as required to reconstruct a visualized image.
The variational self-coding technology in the step a is to acquire a snap-shot hidden trouble picture in the range of the power transmission line and utilize a variational self-coding technology VAE, and the whole structure of the technology is shown in figure 1. This model employs an encoder-decoder framework as shown in fig. 2. The VAE's target constructs a model that generates target data X from hidden variables Z. Assuming that Z obeys some common distribution (such as a normal distribution or a uniform distribution), then a model X ═ g (Z) is trained, which can map the original probability distribution to the probability distributions of the training set to perform the transformation between the distributions.
Assuming that Z follows a normal distribution of criteria, several Zs can be sampled from it1.Z2,...,ZnThen it is transformed to obtain X1＝g(Z1),X2＝g(Z2),...,Xn＝g(Zn)。
First, there is a batch of data samples { X }1,X2,...,XnIs described in its entirety by X, according to { X }1,X2,...,XnObtaining a distribution p (X) of X, p (X) Z ∑ p (X | Z) p (Z), and sampling according to p (X) to obtain all possible xs (including { X ∑ X) }1,X2,...,XnOther than). In this case, p (X | Z) describes a model for generating X from Z, which is implemented by decoder in the network, and assuming Z follows a normal distribution, one can first sample one Z in the normal distribution and then calculate one X from Z. Then, a re-parameterization mechanism is adopted, after the mean value and the variance are calculated by using a neural network, the corresponding normal distribution is not directly sampled, but epsilon is sampled from the standard normal distribution, and then the mean value is multiplied by the variance.
Preferably, step b specifically comprises:
b 1: based on the pictures in the range of the power transmission line in the step a, acquiring hidden danger picture codes by using a variational self-coding technology;
b 2: transmitting the picture codes to a server side through a 4G network;
b 3: collecting the scene hidden danger picture codes of the power transmission line at a server side, and utilizing the obtained scene hidden danger picture codes of the power transmission line;
b 4: and according to the user requirements, reconstructing the codes stored in the server into pictures.
B, transmitting the codes of the captured pictures of the power transmission line to a server through a 4G network, collecting the codes of the pictures of the scene hidden danger of the power transmission line at the server, utilizing the obtained codes of the pictures of the scene hidden danger of the power transmission line, and reconstructing the pictures of the scene hidden danger of the power transmission line at the server, namely, the codes of the pictures obtained in the step a are transmitted into the server through the 4G network, and a generator decoder is utilized at the server to generate new samples which are similar to but different from training data.
Preferably, b2 is specifically: the camera system comprises an access and control device of a 4G network, the control device automatically uploads a code obtained by decoding a monitoring picture captured by a camera to a server, the access device of the 4G network can be connected to a wireless network of a telecommunication operator to remotely transmit the code, and the generated code is transmitted to a background server through the 4G network.
Preferably, the b4 specifically comprises: the stored power line image is encoded as input to a decoder to generate new samples that are similar but different from the training data.
The working process is as follows: firstly, a monitoring camera captures hidden danger pictures in a power transmission line range, picture coding is carried out by using a variational self-coding technology, codes are uploaded to a server through a 4G network, power transmission line scene hidden danger picture codes are collected at the server, the obtained power transmission line scene hidden danger picture codes are utilized, and power transmission line scene hidden danger pictures needing to be displayed are reconstructed at the server according to requirements. The method is based on the variational self-coding technology, and the existing picture with the scene hidden danger of the power transmission line is used for decoding the picture. In the field of intelligent inspection of power transmission channels, through repeated experimental demonstration, 1000-time image lossy compression can be realized after the coding technology is adopted in the power transmission channel image data set of the national power grid, and meanwhile, an image with a good visualization effect can be reconstructed. The method and the device realize the quick and low-cost transmission and storage of the monitoring images of the power transmission line, save the transmission bandwidth and the stored server cost, and are beneficial to realizing the intelligent management of the operation of the power grid.
Claims (5)
1. A power transmission line image decoding method based on variational self-coding is characterized in that: comprises the following steps of (a) carrying out,
a. the method comprises the following steps that a camera captures pictures in a scene range of the power transmission line, and the pictures are coded by using a variational self-coding technology;
b. and transmitting the data obtained by coding to a server through a 4G network, and reconstructing the codes stored in the server into pictures according to requirements.
2. The method according to claim 1, wherein the variational self-coding technique in step a is based on an encoder-decoder mode generation model, a neural network is used in the mobile terminal device to encode the picture, and the encoded data is decoded in the server as required to reconstruct a visualized image.
3. The method for decoding the image of the power transmission line based on the variational self-coding according to claim 1, wherein the step b specifically comprises:
b 1: based on the pictures in the range of the power transmission line in the step a, acquiring hidden danger picture codes by using a variational self-coding technology;
b 2: transmitting the picture codes to a server side through a 4G network;
b 3: collecting the scene hidden danger picture codes of the power transmission line at a server side, and utilizing the obtained scene hidden danger picture codes of the power transmission line;
b 4: and according to the user requirements, reconstructing the codes stored in the server into pictures.
4. The method for decoding the image of the power transmission line based on the variational self-coding according to claim 1, wherein the b2 specifically comprises: the camera system comprises an access and control device of a 4G network, the control device automatically uploads a code obtained by decoding a monitoring picture captured by a camera to a server, the access device of the 4G network can be connected to a wireless network of a telecommunication operator to remotely transmit the code, and the generated code is transmitted to a background server through the 4G network.
5. The method according to claim 1, wherein the b4 specifically comprises: the stored power line image is encoded as input to a decoder to generate new samples that are similar but different from the training data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911101962.5A CN110730352A (en) | 2019-11-12 | 2019-11-12 | Power transmission line image decoding method based on variational self-coding |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911101962.5A CN110730352A (en) | 2019-11-12 | 2019-11-12 | Power transmission line image decoding method based on variational self-coding |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110730352A true CN110730352A (en) | 2020-01-24 |
Family
ID=69223997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911101962.5A Pending CN110730352A (en) | 2019-11-12 | 2019-11-12 | Power transmission line image decoding method based on variational self-coding |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110730352A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113256744A (en) * | 2020-02-10 | 2021-08-13 | 武汉Tcl集团工业研究院有限公司 | Image coding and decoding method and system |
| CN113747178A (en) * | 2021-09-03 | 2021-12-03 | 中科方寸知微(南京)科技有限公司 | Image edge end compression and back end recovery method and system in power channel visualization scene |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010206463A (en) * | 2009-03-03 | 2010-09-16 | Victor Co Of Japan Ltd | Method, device and program for encoding moving image |
| CN105163121A (en) * | 2015-08-24 | 2015-12-16 | 西安电子科技大学 | Large-compression-ratio satellite remote sensing image compression method based on deep self-encoding network |
| CN105447530A (en) * | 2016-01-05 | 2016-03-30 | 国网四川省电力公司检修公司 | Power transmission line hidden risk and fault detection method based on image identification technology |
| CN106657930A (en) * | 2017-02-21 | 2017-05-10 | 山东信通电子股份有限公司 | Image monitoring device and method of power transmission line |
| CN207234978U (en) * | 2017-08-02 | 2018-04-13 | 宁夏龙源电力有限公司 | Intelligent patrol device based on picture intellectual analysis transmission line of electricity |
-
2019
- 2019-11-12 CN CN201911101962.5A patent/CN110730352A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010206463A (en) * | 2009-03-03 | 2010-09-16 | Victor Co Of Japan Ltd | Method, device and program for encoding moving image |
| CN105163121A (en) * | 2015-08-24 | 2015-12-16 | 西安电子科技大学 | Large-compression-ratio satellite remote sensing image compression method based on deep self-encoding network |
| CN105447530A (en) * | 2016-01-05 | 2016-03-30 | 国网四川省电力公司检修公司 | Power transmission line hidden risk and fault detection method based on image identification technology |
| CN106657930A (en) * | 2017-02-21 | 2017-05-10 | 山东信通电子股份有限公司 | Image monitoring device and method of power transmission line |
| CN207234978U (en) * | 2017-08-02 | 2018-04-13 | 宁夏龙源电力有限公司 | Intelligent patrol device based on picture intellectual analysis transmission line of electricity |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113256744A (en) * | 2020-02-10 | 2021-08-13 | 武汉Tcl集团工业研究院有限公司 | Image coding and decoding method and system |
| CN113747178A (en) * | 2021-09-03 | 2021-12-03 | 中科方寸知微(南京)科技有限公司 | Image edge end compression and back end recovery method and system in power channel visualization scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102170554B (en) | Based on network camera device and the network shooting method of power network | |
| CN110099280B (en) | Video service quality enhancement method under limitation of wireless self-organizing network bandwidth | |
| CN110730352A (en) | Power transmission line image decoding method based on variational self-coding | |
| CN109903351B (en) | Image compression method based on combination of convolutional neural network and traditional coding | |
| CN105262992A (en) | Transformer station equipment state analysis inspection tour system based on two-dimension code recognition | |
| CN107547898A (en) | A kind of controllable two-parameter distribution system of sensor of energy consumption precision | |
| CN107657601A (en) | A kind of substation equipment division physical location monitoring device and method | |
| CN116600132A (en) | Coal mine video data self-adaptive compression method | |
| CN112203098A (en) | Mobile terminal image compression method based on edge feature fusion and super-resolution | |
| CN114494569A (en) | Cloud rendering method and device based on lightweight neural network and residual streaming transmission | |
| CN113822954A (en) | Deep learning image coding method for man-machine cooperation scene under resource constraint | |
| CN105025259A (en) | Intelligent transformer station inspection video live broadcast system and method based on Google eyeglasses | |
| CN203482319U (en) | Video monitoring and analysis system for power station workshops | |
| CN112887679A (en) | Lossless video remote acquisition method and system | |
| CN111866509A (en) | Method for realizing real-time video data transmission based on forward error correction and receiving feedback | |
| CN206712926U (en) | A kind of more exposure image emerging systems | |
| CN111479286A (en) | Data processing method for reducing communication flow of edge computing system | |
| CN110418108A (en) | Unattended area monitoring system based on CS theory | |
| CN105306941B (en) | A kind of method for video coding | |
| CN115209147A (en) | Camera video transmission bandwidth optimization method, device, equipment and storage medium | |
| CN103167282B (en) | A kind of online updating method of image compression dynamic regulation | |
| CN112565178A (en) | Unmanned aerial vehicle power equipment system of patrolling and examining based on streaming media technique | |
| CN204669533U (en) | A kind of airborne dust monitoring information acquisition based on compressed sensing and transmission hardware circuit | |
| CN111526157A (en) | Internet of things data acquisition system and method | |
| CN105578192A (en) | Power visual metamodel agglomeration compression method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200124 |