CN107194396A - Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system - Google Patents
Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system Download PDFInfo
- Publication number
- CN107194396A CN107194396A CN201710318998.3A CN201710318998A CN107194396A CN 107194396 A CN107194396 A CN 107194396A CN 201710318998 A CN201710318998 A CN 201710318998A CN 107194396 A CN107194396 A CN 107194396A
- Authority
- CN
- China
- Prior art keywords
- mrow
- early warning
- land resources
- against regulations
- monitoring system
- 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
- 230000033228 biological regulation Effects 0.000 title claims abstract description 34
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 25
- 238000001514 detection method Methods 0.000 claims abstract description 14
- 230000008569 process Effects 0.000 claims abstract description 7
- 238000000889 atomisation Methods 0.000 claims abstract description 6
- 238000012549 training Methods 0.000 claims description 19
- 238000010845 search algorithm Methods 0.000 claims description 5
- 239000000203 mixture Substances 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 3
- 230000001629 suppression Effects 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 claims description 2
- 238000013461 design Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 claims description 2
- 238000011478 gradient descent method Methods 0.000 claims description 2
- 239000003595 mist Substances 0.000 claims description 2
- 238000005457 optimization Methods 0.000 claims description 2
- 230000007704 transition Effects 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 2
- 238000012795 verification Methods 0.000 abstract description 2
- 230000002159 abnormal effect Effects 0.000 abstract 1
- 230000000694 effects Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 208000003464 asthenopia Diseases 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
Method for early warning is recognized based on the specific architecture against regulations in a kind of land resources video monitoring system that the present invention is provided, video camera is set up by control point in the wild, fixed point, periodically timing capture picture and be back to information centre's server.The present invention utilizes the R CNN based on convolutional neural networks on the basis of traditional land resources monitoring network（Fast Region Convolutional Neural Networks）Framework carries out target detection to information centre's server for building.Recognition result in inherent picture of continuous a period of time is carried out to determine whether for the architecture against regulations, most there are abnormal submission manual examination and verification to reach the purpose of intelligent, automatically-monitored land resources at last than analysis.The present invention has been done to the candid photograph picture of passback to go atomization process to reduce influence of the haze weather to testing result based on dark.
Description
Technical field
It is applied to the present invention relates to Computer Applied Technology field, more particularly to one kind in land resources video monitoring system
To the identification method for early warning of the specific architecture against regulations.
Technical background
As a kind of sector application of existing ripe Network Video Surveillance technology, the video monitoring system of land resources is general
It also contains head end video information gathering, Video coding network transmission, the operation management of information centre and data processing, its system
As shown in Figure 1.In actual project construction, increase rapidly with the quantity of control point, stared at for manually on duty by face
The video of the amount of bordering on the sea captures image information, and supervisor faces large number of monitoring image, and to neglect omission unavoidably important
Information, visual fatigue can also cause supervision effect to be not so good as people's will.
Intelligent monitoring based on computer assisted image processing is the developing direction of future video monitoring, can early be found, early pre-
It is alert, while the unattended of large-range monitoring can be realized.For the video monitoring system of land resources, video monitoring and figure
One of major function as capturing system seek to carry out some illegal activities occurred in monitoring range or the architecture against regulations and
When discovery and early warning, be easy to staff to make corresponding processing in time.The Key technique problem being reacted in system is exactly
The identification that the image captured carries out certain objects is put to specifically monitored.Object identification based on image procossing, which is substantially all, to be made
With machine learning method.In order to improve the performance of these methods, it usually needs collect bigger data set, and build strong
Model and prevent over-fitting.But the same type objects in actual environment can generally also show the form of complexity, therefore
The recognition effect reached must just use huge and accurate training set.
Convolutional neural networks CNN is classical in deep learning and one of conventional model, R-CNN (Regions with
Convolutional Neural Network Features) it is a kind of classics in the object detecting areas based on deep learning
Algorithm, Alexnet network models have been one of most successful convolutional neural networks models since 2012, and it is disclosed
Excellent Detection results are achieved on PASCOL_VOC data sets.It is both fresh and unusual comprising some in Alexnet
Feature, they improve the performance of network, and reduce the training time of network.Alexnet networks comprising five convolutional layers and
Three full articulamentums (see accompanying drawing 2).
The content of the invention
The purpose of the present invention is to be based on for application of the existing video monitoring system in land resources monitoring there is provided one kind
The method of quick R-CNN object identifications, for the specific architecture against regulations of early detection in land resources video monitoring system, so that
Reach the effect of timely early warning.
The technical solution adopted in the present invention is：Known in a kind of land resources video monitoring system based on the specific architecture against regulations
Other method for early warning, it is characterised in that comprise the following steps：
Step 1：Based on land resources monitor video, to each video image, division needs region and the record of early warning
The coordinate (m, n) in the region upper left corner and the coordinate (x, y) in the lower right corner, division are needed the region of early warning be used as detection sample；
Step 2：Structure includes the training set of some building pictures, and Alexnet convolutional Neural nets are based on using training set
CNN model of the network training on building, and train Softmax graders using the output for training set of CNN models；
Step 3：Detection sample is carried out going atomization process；
Step 4：Several candidate regions are extracted in detection sample using Selective search algorithms, will each be waited
Favored area normalizes to M × M × L, and wherein M span is 100-250, and L is R/G/B triple channels；Then instruct in step 2
Forward-propagating in the CNN models perfected, extracts the characteristic vector of last layer；
Step 5：The characteristic vector extracted in step 4 is beaten using the Softmax graders trained in step 2
Point, the fraction S of the candidate region extracted in step 4 for building is obtained, if if fraction S is more than threshold value T, marking the time
Favored area is building；
Step 6：The candidate region marked is removed using non-maxima suppression NMS methods and intersects unnecessary frame, if one section
The continuous time candidate region it is unmarked go out building and photo current detects building, then judge the building as the architecture against regulations and pre-
It is alert.
Relative to prior art, the solution have the advantages that：
1. compared to traditional Network Video Surveillance, the present invention can effectively reduce the expense that network traffics are brought.
The present invention realizes automation, intelligentized architecture against regulations early warning simultaneously, and the unattended of large-range monitoring can be achieved；
2. the present invention has done the processing for going atomization to the picture of passback, it can effectively increase the accuracy rate of identification.
Brief description of the drawings
Fig. 1 is the application system framework model of the embodiment of the present invention；
Fig. 2 is Alexnet network architecture models in the embodiment of the present invention；
Fig. 3 is flow chart of the invention.
Embodiment
Understand for the ease of those of ordinary skill in the art and implement the present invention, below in conjunction with the accompanying drawings and embodiment is to this hair
It is bright to be described in further detail, it will be appreciated that implementation example described herein is merely to illustrate and explain the present invention, not
For limiting the present invention.
The present invention be for territory video frequency monitoring system, use specific IMAQ strategy obtain corresponding pictures with
Afterwards, then based on quick R-CNN architecture against regulations model is trained, Building recognition is done to each control point, if continuous a period of time should
Without building, current photo identifies building in area, then judges the building as the architecture against regulations.
A kind of land resources monitoring and early warning method based on quick R-CNN object identifications, by front end to monitoring head
Set, daily in fixed angle, picture is carried out in default particular point in time with fixed camera parameter and captures passback.
After continuous acquisition fixed point captures picture, system will carry out the object identification based on following algorithm for this sequence of pictures,
The picture that the architecture against regulations may be had by being found out in all sequence of pictures is marked, and is exported as system, forms early warning foundation, simultaneously
The architecture against regulations of mark is added into training set.
Input：In some fixed preset position to be identified, in predetermined point of time (such as 6:00、14:00、22:00) grab successively
Clapping picture passback (can also record one section of video, then to video flowing frequency in certain intervals, such as every 5 frames, one frame of interception, enter
Row interception, and packet foundation is used as to unique No. ID of the source distribution of each front end control point, while being capped timestamp with area
Divide the frame of video group of different time node under same ID), input the picture of passback.
Output：The architecture against regulations in picture is recognized on the CNN networks trained, and is marked as the defeated of algorithm
Go out.
It is pre- based on the identification of the specific architecture against regulations in a kind of land resources video monitoring system provided see Fig. 3, the present invention
Alarm method, comprises the following steps：
Step 1：Based on land resources monitor video, to each video image, division needs region and the record of early warning
The coordinate (m, n) in the region upper left corner and the coordinate (x, y) in the lower right corner, division are needed the region of early warning be used as detection sample；
Step 2：Building includes the training set of 1000-2000 buildings, and Alexnet convolutional Neurals are based on using training set
CNN model of the network training on building, and train Softmax graders using the output for training set of CNN models；
Implement including following sub-step：
Step 2.1：The CNN network structure design phases：Using the Alexnet network architectures, the Alexnet network architectures are shown in accompanying drawing
2。
Step 2.2：CNN networks have the supervision pre-training stage：Directly with the good Alexnet of ImageNet database trainings
Netinit parameter, then carries out the small parameter perturbations for having supervision in the next step fine-tuing stages.
Step 2.3：The fine-tuning stages；
1000-2000 pictures are chosen as training set, the upper left side of the architecture against regulations and lower right in picture is marked and sits
Mark, is substituted for architecture against regulations output member and 1 background by last layer of Alexnet convolutional neural networks, uses Selective
Several candidate regions are extracted in the detection sample that search algorithms are divided in step 1, if the candidate region extracted and mark
The coordinate degree of overlapping of note is more than threshold value T, then is otherwise negative sample labeled as positive sample；
Step 2.4：The pre-selection frame packet (batch-size) that 1000-2000 pictures are selected, input to Alexnt
In convolutional neural networks network, stochastic gradient descent method is used when the network optimization is solved, learning rate size is 0.001.
Wherein, batch-size value is 120, and wherein positive sample is 90, and negative sample is 30.
Step 3：Detection sample is carried out going atomization process；
The picture of required detection is photographing outdoors, therefore is easily influenceed by haze weather.Defogging algorithm based on dark is public
Formula is as follows：
Wherein J (x) is the picture after defogging, and I (x) is artwork, and A is global atmosphere light composition, and t (x) is transmissivity, t0For
Threshold value, when t (x) is less than t0When prevent J (x) excessive and cause image to white field transition；
Ω (x) represents a window centered on pixel X, and ω is modelling factors to simulate normal weather situation, its institute
What is calculated is R/G/B triple channels.Parameter ω value is 0.90, parameter t0Value be 0.15.
Global atmosphere light composition A process of asking for is to take preceding 0.2% according to the size of brightness first from dark channel diagram
Pixel；Then this 0.2% pixel it is original have the point with maximum brightness is found in mist video image I (x) correspondence positions
Value, be used as A values.
Step 4：Several candidate regions are extracted in detection sample using Selective search algorithms, will each be waited
Favored area normalizes to M × M × L；Then forward-propagating in the CNN models trained in step 2, the spy for extracting last layer
Levy vector.Wherein M span is about 100-250, and different sizes needs corresponding in corresponding modification Alexnet networks
Parameter, L is R/G/B triple channels.
Step 5：The characteristic vector extracted in step 4 is beaten using the Softmax graders trained in step 2
Point, the fraction S of the candidate region extracted in step 4 for building is obtained, if if fraction S is more than some threshold value T, marking
The candidate region is building；
Step 6：The candidate region marked in step 5 is removed using non-maxima suppression (NMS) method and intersects unnecessary frame.
If in the monitor area divided in one section of continuous time step 1 it is unmarked go out building and photo current detect building, judge
The building is the architecture against regulations and early warning and submits manual examination and verification.If the architecture against regulations, then architecture against regulations training set is added.
All picture arrays of certain website to obtaining carry out feature recognition processing respectively, and return to recognition result as pre-
Alert foundation；The algorithm flow of identification is as shown in Figure 3.
The method that the present invention is designed is suitable for the monitoring system application scenarios of large scale deployment.The effect monitored on a large scale is only
Judge by eye-observation video, not only inefficiency, and produce little effect.Method designed by the present invention passes through to passing back
Sequence of pictures carries out the intelligent business demand for comparing analysis, being monitored for territory, and picture will be returned by Intelligent Recognition algorithm
In the doubtful architecture against regulations be labeled, realize early warning, will be freed in the pictorial information of magnanimity in administrative staff, greatly
The big operation efficiency for improving system.
It should be appreciated that the part that this specification is not elaborated belongs to prior art.
It should be appreciated that the above-mentioned description for preferred embodiment is more detailed, therefore it can not be considered to this
The limitation of invention patent protection scope, one of ordinary skill in the art is not departing from power of the present invention under the enlightenment of the present invention
Profit is required under protected ambit, can also be made replacement or be deformed, each fall within protection scope of the present invention, this hair
It is bright scope is claimed to be determined by the appended claims.
Claims (9)
1. method for early warning is recognized based on the specific architecture against regulations in a kind of land resources video monitoring system, it is characterised in that including
Following steps：
Step 1：Based on land resources monitor video, to each video image, the region for needing early warning and posting field are divided
The coordinate (m, n) in the upper left corner and the coordinate (x, y) in the lower right corner, division are needed the region of early warning be used as detection sample；
Step 2：Structure includes the training set of some building pictures, is instructed using training set based on Alexnet convolutional neural networks
Practice the CNN models on building, and Softmax graders are trained using the output for training set of CNN models；
Step 3：Detection sample is carried out going atomization process；
Step 4：Several candidate regions are extracted in detection sample using Selective search algorithms, by each candidate regions
Domain normalizes to M × M × L, and wherein M span is 100-250, and L is R/G/B triple channels；Then train in step 2
CNN models in forward-propagating, extract the characteristic vector of last layer；
Step 5：The characteristic vector extracted in step 4 is given a mark using the Softmax graders trained in step 2, obtained
The candidate region extracted into step 4 for building fraction S, if fraction S be more than threshold value T if, mark the candidate regions
Domain is building；
Step 6：The candidate region marked is removed using non-maxima suppression NMS methods and intersects unnecessary frame, if one section continuous
The time candidate region it is unmarked go out building and photo current detects building, then judge the building as the architecture against regulations and early warning.
2. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 1,
Characterized in that, step 2 is implemented including following sub-step：
Step 2.1：Use Alexnet network architecture design CNN network structures；
Step 2.2：Use Alexnet netinit parameter；
Step 2.3：1000-2000 pictures are chosen as training set, the upper left side and bottom right of the architecture against regulations in picture is marked
Square coordinate, is substituted for architecture against regulations output member and 1 background by last layer of Alexnet convolutional neural networks, uses afterwards
Selective search algorithms extract candidate frame, if the coordinate degree of overlapping of the candidate frame extracted and mark is more than threshold value T,
It is otherwise negative sample labeled as positive sample；
Step 2.4：The candidate frame extracted in step 2.3 is grouped and inputted into Alexnet convolutional neural networks networks, network
Stochastic gradient descent method is used during Optimization Solution, learning rate size is 0.001.
3. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 2,
It is characterized in that：In step 2.4, candidate frame is grouped into 120 groups, and wherein positive sample is 90 groups, and negative sample is 30 groups.
4. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 1,
Characterized in that, in step 3, the atomization process formula that goes based on dark is：
<mrow>
<mi>J</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>A</mi>
</mrow>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
<mo>,</mo>
<msub>
<mi>t</mi>
<mn>0</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>+</mo>
<mi>A</mi>
<mo>;</mo>
</mrow>
Wherein J (x) is the video image after defogging, and I (x) is original video image, and A is global atmosphere light composition, and t (x) is transmission
Rate, t0For threshold value, when t (x) is less than t0When prevent J (x) excessive and cause image to white field transition；
<mrow>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mn>1</mn>
<mo>-</mo>
<msub>
<mi>&omega;min</mi>
<mrow>
<mi>y</mi>
<mo>&Element;</mo>
<mi>&Omega;</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
<mfrac>
<mrow>
<mi>I</mi>
<mrow>
<mo>(</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
<mi>A</mi>
</mfrac>
<mo>)</mo>
</mrow>
<mo>;</mo>
</mrow>
Ω (x) represents a window centered on pixel X, and ω is modelling factors, and for simulating normal weather situation, it is counted
What is calculated is R/G/B triple channels.
5. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 4,
Characterized in that, in step 3, ω value is 0.90.
6. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 4,
Characterized in that, in step 3, t0Value be 0.15.
7. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 4,
It is characterized in that：The composition of global atmosphere light described in step 3 A process of asking for is, first according to brightness from dark channel diagram
Size takes preceding 0.2% pixel, then this 0.2% pixel it is original have tool is found in mist video image I (x) correspondence positions
There is the value of the point of maximum brightness, be used as A values.
8. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 1,
Characterized in that, in step 5, fraction S values are 0.6.
9. method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system according to claim 1,
Characterized in that, in step 5, threshold value T value is 0.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710318998.3A CN107194396A (en) | 2017-05-08 | 2017-05-08 | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710318998.3A CN107194396A (en) | 2017-05-08 | 2017-05-08 | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107194396A true CN107194396A (en) | 2017-09-22 |
Family
ID=59873816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710318998.3A Pending CN107194396A (en) | 2017-05-08 | 2017-05-08 | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107194396A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109001833A (en) * | 2018-06-22 | 2018-12-14 | 天和防务技术（北京）有限公司 | A kind of Terahertz hazardous material detection method based on deep learning |
| CN109166091A (en) * | 2018-06-29 | 2019-01-08 | 中国移动通信集团江苏有限公司 | Detection method, device, equipment and medium |
| CN109190608A (en) * | 2018-10-30 | 2019-01-11 | 长威信息科技发展股份有限公司 | A kind of city intelligent identification Method violating the regulations |
| CN109492753A (en) * | 2018-11-05 | 2019-03-19 | 中山大学 | A kind of method of the stochastic gradient descent of decentralization |
| CN109670515A (en) * | 2018-12-13 | 2019-04-23 | 南京工业大学 | Method and system for detecting building change in unmanned aerial vehicle image |
| CN110096943A (en) * | 2019-01-28 | 2019-08-06 | 浙江浩腾电子科技股份有限公司 | A kind of architecture against regulations detection system based on deep learning |
| CN110399820A (en) * | 2019-07-16 | 2019-11-01 | 西安工程大学 | A kind of margin of roads scenery visual identity analysis method |
| CN110599544A (en) * | 2019-08-08 | 2019-12-20 | 佛山科学技术学院 | Workpiece positioning method and device based on machine vision |
| CN111339353A (en) * | 2020-02-21 | 2020-06-26 | 北京容联易通信息技术有限公司 | Image processing self-optimization method and device, electronic equipment and storage medium |
| CN111737285A (en) * | 2020-05-12 | 2020-10-02 | 深圳震有科技股份有限公司 | Building query and annotation processing method and device based on geospatial analysis |
| CN112241659A (en) * | 2019-07-17 | 2021-01-19 | 杭州海康威视数字技术股份有限公司 | Detection device and detection method for illegal building and terminal equipment |
| CN113052125A (en) * | 2021-04-09 | 2021-06-29 | 内蒙古科电数据服务有限公司 | Construction site violation image recognition and alarm method |
| CN115797775A (en) * | 2022-12-14 | 2023-03-14 | 中国铁塔股份有限公司重庆市分公司 | Intelligent illegal building identification method and system based on near-earth video image |
| CN116563774A (en) * | 2023-02-23 | 2023-08-08 | 中国铁塔股份有限公司重庆市分公司 | Automatic identification method and system for illegal mining of cultivated land in China |
| CN117475296A (en) * | 2023-08-16 | 2024-01-30 | 中国铁塔股份有限公司重庆市分公司 | Real-time monitoring method and system for rural illegal building |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102446390A (en) * | 2011-03-16 | 2012-05-09 | 上海市电力公司 | Method and system for safety detection and early warning of monitoring areas near power transmission lines |
| CN103034860A (en) * | 2012-12-14 | 2013-04-10 | 南京思创信息技术有限公司 | Scale-invariant feature transform (SIFT) based illegal building detection method |
| CN205121797U (en) * | 2015-10-13 | 2016-03-30 | 湖南人文科技学院 | Architecture against regulations automated inspection warning system based on 4G |
| CN105844257A (en) * | 2016-04-11 | 2016-08-10 | 吉林大学 | Early warning system based on machine vision driving-in-fog road denoter missing and early warning method |
| CN105929845A (en) * | 2016-05-18 | 2016-09-07 | 中国计量大学 | Unmanned aerial vehicle network-based river channel cruise system and cruise method |
-
2017
- 2017-05-08 CN CN201710318998.3A patent/CN107194396A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102446390A (en) * | 2011-03-16 | 2012-05-09 | 上海市电力公司 | Method and system for safety detection and early warning of monitoring areas near power transmission lines |
| CN103034860A (en) * | 2012-12-14 | 2013-04-10 | 南京思创信息技术有限公司 | Scale-invariant feature transform (SIFT) based illegal building detection method |
| CN205121797U (en) * | 2015-10-13 | 2016-03-30 | 湖南人文科技学院 | Architecture against regulations automated inspection warning system based on 4G |
| CN105844257A (en) * | 2016-04-11 | 2016-08-10 | 吉林大学 | Early warning system based on machine vision driving-in-fog road denoter missing and early warning method |
| CN105929845A (en) * | 2016-05-18 | 2016-09-07 | 中国计量大学 | Unmanned aerial vehicle network-based river channel cruise system and cruise method |
Non-Patent Citations (2)
| Title |
|---|
| ROSS GIRSHICK 等: "Rich feature hierarchies for accurate object detection and semantic segmentation", 《THE IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION(CVPR)》 * |
| 王万国 等: "基于RCNN的无人机巡检图像电力小部件识别研究", 《地球信息科学》 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109001833A (en) * | 2018-06-22 | 2018-12-14 | 天和防务技术（北京）有限公司 | A kind of Terahertz hazardous material detection method based on deep learning |
| CN109166091B (en) * | 2018-06-29 | 2020-11-06 | 中国移动通信集团江苏有限公司 | Detection method, apparatus, device and medium |
| CN109166091A (en) * | 2018-06-29 | 2019-01-08 | 中国移动通信集团江苏有限公司 | Detection method, device, equipment and medium |
| CN109190608A (en) * | 2018-10-30 | 2019-01-11 | 长威信息科技发展股份有限公司 | A kind of city intelligent identification Method violating the regulations |
| CN109492753A (en) * | 2018-11-05 | 2019-03-19 | 中山大学 | A kind of method of the stochastic gradient descent of decentralization |
| CN109670515A (en) * | 2018-12-13 | 2019-04-23 | 南京工业大学 | Method and system for detecting building change in unmanned aerial vehicle image |
| CN110096943A (en) * | 2019-01-28 | 2019-08-06 | 浙江浩腾电子科技股份有限公司 | A kind of architecture against regulations detection system based on deep learning |
| CN110399820A (en) * | 2019-07-16 | 2019-11-01 | 西安工程大学 | A kind of margin of roads scenery visual identity analysis method |
| CN112241659B (en) * | 2019-07-17 | 2024-05-17 | 杭州海康威视数字技术股份有限公司 | Detection device and detection method for illegal building and terminal equipment |
| CN112241659A (en) * | 2019-07-17 | 2021-01-19 | 杭州海康威视数字技术股份有限公司 | Detection device and detection method for illegal building and terminal equipment |
| CN110599544B (en) * | 2019-08-08 | 2023-01-24 | 佛山科学技术学院 | Workpiece positioning method and device based on machine vision |
| CN110599544A (en) * | 2019-08-08 | 2019-12-20 | 佛山科学技术学院 | Workpiece positioning method and device based on machine vision |
| CN111339353B (en) * | 2020-02-21 | 2023-06-09 | 北京容联易通信息技术有限公司 | Image processing self-optimization method and device, electronic equipment and storage medium |
| CN111339353A (en) * | 2020-02-21 | 2020-06-26 | 北京容联易通信息技术有限公司 | Image processing self-optimization method and device, electronic equipment and storage medium |
| CN111737285A (en) * | 2020-05-12 | 2020-10-02 | 深圳震有科技股份有限公司 | Building query and annotation processing method and device based on geospatial analysis |
| CN113052125B (en) * | 2021-04-09 | 2022-10-28 | 内蒙古科电数据服务有限公司 | Construction site violation image recognition and alarm method |
| CN113052125A (en) * | 2021-04-09 | 2021-06-29 | 内蒙古科电数据服务有限公司 | Construction site violation image recognition and alarm method |
| CN115797775A (en) * | 2022-12-14 | 2023-03-14 | 中国铁塔股份有限公司重庆市分公司 | Intelligent illegal building identification method and system based on near-earth video image |
| CN115797775B (en) * | 2022-12-14 | 2024-04-26 | 中国铁塔股份有限公司重庆市分公司 | Intelligent illegal building identification method and system based on near-to-ground video image |
| CN116563774A (en) * | 2023-02-23 | 2023-08-08 | 中国铁塔股份有限公司重庆市分公司 | Automatic identification method and system for illegal mining of cultivated land in China |
| CN116563774B (en) * | 2023-02-23 | 2024-08-09 | 中国铁塔股份有限公司重庆市分公司 | Automatic identification method and system for illegal mining of cultivated land in China |
| CN117475296A (en) * | 2023-08-16 | 2024-01-30 | 中国铁塔股份有限公司重庆市分公司 | Real-time monitoring method and system for rural illegal building |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107194396A (en) | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system | |
| CN110807353B (en) | Substation foreign matter identification method, device and system based on deep learning | |
| CN108154134B (en) | Pornographic image detection method is broadcast live in internet based on depth convolutional neural networks | |
| CN111414887B (en) | Secondary detection mask face recognition method based on YOLOV3 algorithm | |
| CN106960195B (en) | Crowd counting method and device based on deep learning | |
| CN110717481B (en) | Method for realizing face detection by using cascaded convolutional neural network | |
| CN110826514A (en) | Construction site violation intelligent identification method based on deep learning | |
| CN110070530A (en) | A kind of powerline ice-covering detection method based on deep neural network | |
| CN110084165A (en) | The intelligent recognition and method for early warning of anomalous event under the open scene of power domain based on edge calculations | |
| CN109376637A (en) | Passenger number statistical system based on video monitoring image processing | |
| CN106097346A (en) | A kind of video fire hazard detection method of self study | |
| CN105208325B (en) | The land resources monitoring and early warning method captured and compare analysis is pinpointed based on image | |
| CN109190608A (en) | A kind of city intelligent identification Method violating the regulations | |
| CN113486865A (en) | Power transmission line suspended foreign object target detection method based on deep learning | |
| CN113887412B (en) | Detection method, detection terminal, monitoring system and storage medium for pollution emission | |
| CN106339657B (en) | Crop straw burning monitoring method based on monitor video, device | |
| CN111723656B (en) | Smog detection method and device based on YOLO v3 and self-optimization | |
| CN108229524A (en) | A kind of chimney and condensing tower detection method based on remote sensing images | |
| CN111091110A (en) | Wearing identification method of reflective vest based on artificial intelligence | |
| CN109389146A (en) | Equipment state feedback method and system based on neural network image sorting algorithm | |
| CN115294519A (en) | Abnormal event detection and early warning method based on lightweight network | |
| CN108664886A (en) | A kind of fast face recognition method adapting to substation's disengaging monitoring demand | |
| CN116884192A (en) | Power production operation risk early warning method, system and equipment | |
| CN109902730A (en) | Broken strand of power transmission line detection method based on deep learning | |
| CN109450876A (en) | A kind of DDos recognition methods and system based on various dimensions state-transition matrix feature |
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: 20170922 |