CN106991668A - A kind of evaluation method of day net camera shooting picture - Google Patents
A kind of evaluation method of day net camera shooting picture Download PDFInfo
- Publication number
- CN106991668A CN106991668A CN201710136626.9A CN201710136626A CN106991668A CN 106991668 A CN106991668 A CN 106991668A CN 201710136626 A CN201710136626 A CN 201710136626A CN 106991668 A CN106991668 A CN 106991668A
- Authority
- CN
- China
- Prior art keywords
- picture
- vehicle
- face
- camera shooting
- motion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
- G06F18/2193—Validation; Performance evaluation; Active pattern learning techniques based on specific statistical tests
-
- 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
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
-
- 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
- G06T2207/30168—Image quality inspection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N17/00—Diagnosis, testing or measuring for television systems or their details
- H04N2017/008—Diagnosis, testing or measuring for television systems or their details for television teletext
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses the evaluation method of a kind of day net camera shooting picture, belong to image quality evaluation field.Comprise the following steps：1. the picture photographed of pair camera carries out motion detection, the motion hot spot region of foundation calculates the ratio of moving region, cutting image；2. inputting original image, the moving region having been detected by is inputted into the good convolutional neural networks of training in advance, the target identified is exported and classify；3. by training mutually isostructural convolutional neural networks positioning licence plate or face；4. by ratio shared by motion target area, the quantity of vehicle after positioning and the car plate and face of pedestrian's quantity and identification is evaluated the effect of day net camera shooting picture.The evaluation method can realize the quality for automatically analyzing camera shooting picture, only need to input camera collection video just it is exportable it is corresponding evaluate camera parameter, substantial amounts of manpower and materials can be saved.
Description
Technical field
The invention belongs to image quality evaluation field, more specifically to a kind of day net camera shooting picture
Evaluation method.
Background technology
" day net " is the demand for meeting urban society's the system for crime prevention and control construction and integrated management, with video image information
Obtain, transmit, control, equipment and the related software such as display are regarded to the city that designated area carries out monitoring and information gathering in real time
Frequency monitoring system.Building up for system can realize overall control for social security integrated management, quick response, and strike is accurate to be provided
The technological meanses of modernization.
But during the construction and use of video surveillance network, some problems are also exposed, as monitored overstocked make of layouting
Into wasting, it is laid out and unreasonable causes to omit and dead angle, camera angle are unreasonable causes monitoring not in place etc., the limitation of these problems
The shooting effect and quality of day net camera.And in face of having a very wide distribution, the day net camera of substantial amounts, it is ensured that every
Rationally, its maintenance work will be very difficult for the clear, normal of one picture monitoring, angle.
Current known, day net camera shooting effect also relies on the subjective judgement of people, the angle or parameter of camera
Also need manually to adjust.Therefore intelligent image treatment technology and convolutional neural networks are combined, research passes through computer testing point
Analysis system carries out automatic detection to camera shooting effect and makes evaluation, can by the evaluation to camera shooting picture quality
Suggestion is provided to monitor cloth in the construction for video surveillance network, while can prevent camera shooting picture is ropy to ask
Topic, this has reformed into a research tendency at this stage.
The content of the invention
For also needing in the prior art by manually analyzing the problem of expending a large amount of manpower and materials, the invention provides
The evaluation method of a kind of day net camera quality, it can realize the objective parameter of acquisition, and mesh is moved in such as effective monitoring region
Distinguishable degree of target etc., automatically analyzes the quality of camera shooting picture, passes through the evaluation to camera shooting picture quality
Suggestion can be provided to monitor cloth in the construction of video surveillance network, while can prevent camera shooting picture is ropy to ask
Topic, so as to save a large amount of manpower and materials.
The purpose of the present invention is achieved through the following technical solutions.
The evaluation method of its net camera shooting picture, it is characterised in that comprise the following steps：
S1. motion detection is carried out to the picture photographed of camera, the motion hot spot region of foundation calculates moving region
Ratio, cutting image；
S2. original image is inputted, the moving region having been detected by is inputted into the good vehicle of training in advance, pedestrian recognizes net
Network, exports the target identified, by the target training network of preliminary making, after the completion of training, by what is detected in S1
Moving region input vehicle, pedestrian's identification convolutional network carry out the identification and positioning of target；
S3. similar S2, by training mutually isostructural car plate, recognition of face network positions car plate or face, inputs and is
The vehicle of positioning and the picture of pedestrian, are output as the car plate and face of positioning；During detection, by the vehicle obtained in S2 or pedestrian
Picture carry out the identification of face or car plate；
S4. by ratio shared by motion target area, vehicle after positioning and pedestrian's quantity and the car plate of identification and
The quantity of face is evaluated the effect of day net camera shooting picture.
Further, motion detection uses frame difference method in the step S1, extracts partial frame in video, orients motion
Hot spot region, calculates the maximum boundary rectangle of motion hot spot region, and is used as following neutral net according to this rectangle cutting picture
Input.
Further, it is collection day net camera that same vehicle, recognition of face network step are trained in the step S3
The position of vehicle in the picture of shooting, hand labeled picture, using scaling at random and changing original image size, while color in HVS
In the colour space, the exposure and saturation degree of image are adjusted at random with 1.5 times of the factor.
Further, it is 3000 cars that the data set in same vehicle, recognition of face network is trained in the step S3
Picture and 3000 pedestrian's pictures, pass through manual mark car plate or face.
Compared to prior art, the advantage of the invention is that：
(1) the inventive method is based entirely on the method for objectively evaluating of algorithm, it is adaptable to any camera, only need to input shooting
Head collection video just it is exportable it is corresponding output evaluate camera parameter；
(2) the inventive method can be run in terminal, can save substantial amounts of manpower and materials；
(3) present invention in evaluation quality parameter include shoot effective coverage ratio, the distinguishable degree of photographic subjects and
The recognizable degree of photographic subjects；Effective coverage ratio is motion hot spot region and the ratio of whole image, target it is recognizable
Degree is the vehicles or pedestrians ratio that can be recognized in image, and the recognizable degree of target is the face or car plate that can be positioned
Ratio, three parameters more can comprehensively evaluate the quality of camera shooting.
Brief description of the drawings
Fig. 1 is overall structure diagram of the invention；
Fig. 2 is traffic locating effect figure；
Fig. 3 is License Plate design sketch；
Fig. 4 is the network that moving region is detected and target is positioned；
Embodiment
With reference to Figure of description and specific embodiment, the present invention is described in detail.
Embodiment 1
As shown in figure 1, the inventive method includes following several steps：
1) motion detection is carried out to the picture that photographs of camera first, due to the background picture of shooting be it is static,
So ripe moving target detecting method can directly be used, such as：Frame differential method, single Gaussian Background model, mixing are high
This model etc..Gauss hybrid models are distributed a kind of mould of background pixel of correspondence by multiple Gaussian Profiles to background modeling, each
State, enables adaptation to the background perturbation problem in pixel aspect, and can be enabled the system to adaptive by the continuous renewal to background
Answer the change of background.But, gauss hybrid models change for global illumination, shade is very sensitive, for slow motion mesh
Mark Detection results also undesirable.There is illumination and the change of shade in the picture that the day net camera of intersection is shot, this
Vehicle or the motion ratio of pedestrian are slower sometimes under scene, and gauss hybrid models effect is unsatisfactory.The master of frame differential method
It is exactly that the region moved is detected using two continuous frames in sequence of video images or the difference of three frames to want principle.Although detection
Profile it is unsatisfactory, but mark moving region enough.So intending carrying out the detection of moving target using frame differential method
And target area is drawn in figure.
2) convolutional neural networks are that developed recently gets up and caused a kind of efficient identification method paid attention to extensively, and the network is kept away
The complicated early stage pretreatment to image is exempted from, original image can be directly inputted, so directly by the motor area having been detected by
The good convolutional neural networks of domain input training in advance, export and classify the target identified.In this study, targeted species
Mainly there are pedestrian, bus and car.Knowing for effective target is calculated by the vehicle or pedestrian's number that have navigated to
Not other rate.
3) by choosing car plate and face training network manually.The vehicle pictures navigated to and pedestrian are schemed in test process
In the convolutional neural networks of piece input pre-training, car plate or the people of positioning are exported if it can navigate to car plate or face
Face, positioning licence plate or people are unable to if image is excessively obscured, and effective mesh is calculated by the car plate or face number that navigate to
The distinguishable rate of target.
4) by the area of the moving region detected, the recognizable rate of effective target, the distinguishable rate pair of effective target
The effect of its net camera is evaluated.
Concretely comprise the following steps：1. moving region is detected.Calculated using frame difference method and hot spot region is moved in video, shared ratio note
For R1, R1 is effective monitoring region, and according to motion hot spot region cutting picture, deletes in picture and be worth less part, defeated
Go out the maximum boundary rectangle based on motion hot spot region.
2. will motion hot spot region input vehicle, pedestrian's identification network.Traffic recognition effect is as shown in Figure 2.Training
The framework used is darknet, and Network Details such as Fig. 3, network includes 24 convolutional layers and 2 full articulamentums, and loss function is
Sum of squared errors function.Learning rate is from 10-3To 10-2It is slowly increased, plan was avoided using dropout and data set extending method
Close, dropout random (rate=0.5) after first layer deletes some hidden layer neurons, so as to avoid phase between layers
Coadaptation.Training set is collected in vehicle or pedestrian in the picture that day net camera is shot, hand labeled picture, by data
After collection enhancing, training set quantity is about 5000 width pictures, wherein each picture includes multiple targets.Count oriented pedestrian
It is n1 with vehicle fleet size, and calculating accounts for distinguishable degree of the ratio R2, R2 of traffic sum n in video for effective target,
R2=n1/n.Vehicle and the picture of pedestrian that network output is navigated to, and it is normalized to 227x227 pixel sizes.
3. the vehicle of positioning and pedestrian's picture are inputted into car plate, Face detection network, structure is similar to the network in 2, instruction
Practice car plate or face that data are similarly hand labeled, training set quantity is 3000 width images, wherein being included in each image
One vehicle or pedestrian.Statistics is capable of the car plate or picture number n2 of fixation and recognition, calculates the recognizable rate of effective target
R3=n1/n2.License Plate design sketch is as shown in Figure 4.
4. finally by analysis R1, R2, R3 numerical value, it can know from effective monitoring region, the distinguishable degree of target, target respectively
The quality of other degree analyzing camera shooting picture, calculates G=R1+R2+R3, can obtain the overall scoring of camera.Specifically
Evaluation reference following table：
| Camera score G | Evaluate |
| G<=1 | Difference |
| 1<G<=2 | In |
| 2<G<=2.5 | It is good |
| 2.5<G | It is excellent |
Schematically the invention and embodiments thereof are described above, the description does not have restricted, accompanying drawing
Shown in be also the invention one of embodiment, actual structure is not limited thereto.So, if this area
Those of ordinary skill enlightened by it, in the case where not departing from this creation objective, designed and the technology without creative
The similar frame mode of scheme and embodiment, all should belong to the protection domain of this patent.
Claims (4)
1. the evaluation method of a kind of day net camera shooting picture, it is characterised in that comprise the following steps：
S1. motion detection is carried out to the picture photographed of camera, the motion hot spot region of foundation calculates the ratio of moving region
Rate and cutting image；
S2. original image is inputted, the moving region having been detected by is inputted into the good vehicle of training in advance, pedestrian recognizes network,
The target identified is exported, by the target training network of preliminary making, after the completion of training, by the motion detected in S1
Region input vehicle, pedestrian's identification convolutional network carry out the identification and positioning of target；
S3. similar S2, by training mutually isostructural car plate, recognition of face network positions car plate or face, is inputted as positioning
Vehicle and pedestrian picture, be output as positioning car plate and face, during detection, by the vehicle obtained in S2 or the figure of pedestrian
Piece carries out the identification of face or car plate；
S4. by ratio shared by motion target area, vehicle after positioning and pedestrian's quantity and the car plate and face of identification
Quantity the effect of day net camera shooting picture is evaluated.
2. the evaluation method of a kind of day net camera shooting picture quality according to claim 1, it is characterised in that described
Motion detection uses frame difference method in step S1, extracts partial frame in video, orients motion hot spot region, calculates motion hot zone
The maximum boundary rectangle in domain, and according to input of this rectangle cutting picture as neutral net.
3. the evaluation method of a kind of day net camera shooting picture quality according to claim 1, it is characterised in that described
The picture for training mutually isostructural vehicle, recognition of face network step to be shot for collection day net camera in step S3, manual mark
Remember the position of vehicle in picture, scale using random and change original image size, while in HVS color spaces, with 1.5 times
The factor adjusts the exposure and saturation degree of image at random.
4. the evaluation method of a kind of day net camera shooting picture quality according to claim 1, it is characterised in that described
The data set that mutually isostructural vehicle, recognition of face network are used is trained in step S3 for 3000 vehicle pictures and 3000 rows
The mark car plate or face of people's picture manually.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710136626.9A CN106991668B (en) | 2017-03-09 | 2017-03-09 | Evaluation method for pictures shot by skynet camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710136626.9A CN106991668B (en) | 2017-03-09 | 2017-03-09 | Evaluation method for pictures shot by skynet camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106991668A true CN106991668A (en) | 2017-07-28 |
| CN106991668B CN106991668B (en) | 2020-08-18 |
Family
ID=59411582
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710136626.9A Active CN106991668B (en) | 2017-03-09 | 2017-03-09 | Evaluation method for pictures shot by skynet camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106991668B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107396094A (en) * | 2017-08-17 | 2017-11-24 | 上海大学 | The automatic testing method of single camera damage towards in multi-cam monitoring system |
| CN107978185A (en) * | 2017-12-07 | 2018-05-01 | 何旭连 | A kind of good children learning machine of teaching efficiency |
| CN109063630A (en) * | 2018-07-27 | 2018-12-21 | 北京以萨技术股份有限公司 | A kind of fast vehicle detection method based on separable convolution technique and frame difference compensation policy |
| CN109492637A (en) * | 2018-11-06 | 2019-03-19 | 西安艾润物联网技术服务有限责任公司 | Method of adjustment, identifier terminal and the readable storage medium storing program for executing of cog region |
| CN109919008A (en) * | 2019-01-23 | 2019-06-21 | 平安科技（深圳）有限公司 | Moving target detecting method, device, computer equipment and storage medium |
| CN110135223A (en) * | 2018-02-08 | 2019-08-16 | 浙江宇视科技有限公司 | Method for detecting human face and device |
| CN110751678A (en) * | 2018-12-12 | 2020-02-04 | 北京嘀嘀无限科技发展有限公司 | Moving object detection method and device and electronic equipment |
| CN111626106A (en) * | 2020-04-17 | 2020-09-04 | 惠州市德赛西威智能交通技术研究院有限公司 | Camera vehicle detection rate statistical method and device |
| CN112862855A (en) * | 2019-11-12 | 2021-05-28 | 北京京邦达贸易有限公司 | Image annotation method and device, computing equipment and storage medium |
| CN113453055A (en) * | 2020-03-25 | 2021-09-28 | 华为技术有限公司 | Method and device for generating video thumbnail and electronic equipment |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3627279B2 (en) * | 1995-03-31 | 2005-03-09 | ソニー株式会社 | Quantization apparatus and quantization method |
| US20140152799A1 (en) * | 2011-06-21 | 2014-06-05 | Hamamatsu Photonics K.K. | Light measurement device, light measurement method, and light measurement program |
| CN103871077A (en) * | 2014-03-06 | 2014-06-18 | 中国人民解放军国防科学技术大学 | Extraction method for key frame in road vehicle monitoring video |
| CN104299419A (en) * | 2014-10-28 | 2015-01-21 | 合肥指南针电子科技有限责任公司 | High-definition snapshot information system and method |
| JP2015095882A (en) * | 2013-11-14 | 2015-05-18 | キヤノン株式会社 | Image processing system, image processing method, and program |
| CN104778446A (en) * | 2015-03-19 | 2015-07-15 | 南京邮电大学 | Method for constructing image quality evaluation and face recognition efficiency relation model |
| CN105447487A (en) * | 2015-08-27 | 2016-03-30 | 中山大学 | Evaluation method and system for vehicle license plate identification system |
| CN106203318A (en) * | 2016-06-29 | 2016-12-07 | 浙江工商大学 | The camera network pedestrian recognition method merged based on multi-level depth characteristic |
| CN106446895A (en) * | 2016-10-28 | 2017-02-22 | 安徽四创电子股份有限公司 | License plate recognition method based on deep convolutional neural network |
-
2017
- 2017-03-09 CN CN201710136626.9A patent/CN106991668B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3627279B2 (en) * | 1995-03-31 | 2005-03-09 | ソニー株式会社 | Quantization apparatus and quantization method |
| US20140152799A1 (en) * | 2011-06-21 | 2014-06-05 | Hamamatsu Photonics K.K. | Light measurement device, light measurement method, and light measurement program |
| JP2015095882A (en) * | 2013-11-14 | 2015-05-18 | キヤノン株式会社 | Image processing system, image processing method, and program |
| CN103871077A (en) * | 2014-03-06 | 2014-06-18 | 中国人民解放军国防科学技术大学 | Extraction method for key frame in road vehicle monitoring video |
| CN104299419A (en) * | 2014-10-28 | 2015-01-21 | 合肥指南针电子科技有限责任公司 | High-definition snapshot information system and method |
| CN104778446A (en) * | 2015-03-19 | 2015-07-15 | 南京邮电大学 | Method for constructing image quality evaluation and face recognition efficiency relation model |
| CN105447487A (en) * | 2015-08-27 | 2016-03-30 | 中山大学 | Evaluation method and system for vehicle license plate identification system |
| CN106203318A (en) * | 2016-06-29 | 2016-12-07 | 浙江工商大学 | The camera network pedestrian recognition method merged based on multi-level depth characteristic |
| CN106446895A (en) * | 2016-10-28 | 2017-02-22 | 安徽四创电子股份有限公司 | License plate recognition method based on deep convolutional neural network |
Non-Patent Citations (1)
| Title |
|---|
| 阮体洪: "基于交通视频的车辆检测和车辆行为识别研究", <<中国优秀硕士学位论文全文数据库 (信息科技辑)>> * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107396094A (en) * | 2017-08-17 | 2017-11-24 | 上海大学 | The automatic testing method of single camera damage towards in multi-cam monitoring system |
| CN107396094B (en) * | 2017-08-17 | 2019-02-22 | 上海大学 | Automatic testing method towards camera single in multi-cam monitoring system damage |
| CN107978185A (en) * | 2017-12-07 | 2018-05-01 | 何旭连 | A kind of good children learning machine of teaching efficiency |
| CN107978185B (en) * | 2017-12-07 | 2019-12-03 | 南京白下高新技术产业园区投资发展有限责任公司 | A kind of good children learning machine of teaching efficiency |
| CN110135223A (en) * | 2018-02-08 | 2019-08-16 | 浙江宇视科技有限公司 | Method for detecting human face and device |
| CN109063630A (en) * | 2018-07-27 | 2018-12-21 | 北京以萨技术股份有限公司 | A kind of fast vehicle detection method based on separable convolution technique and frame difference compensation policy |
| CN109063630B (en) * | 2018-07-27 | 2022-04-26 | 以萨技术股份有限公司 | Rapid vehicle detection method based on separable convolution technology and frame difference compensation strategy |
| CN109492637A (en) * | 2018-11-06 | 2019-03-19 | 西安艾润物联网技术服务有限责任公司 | Method of adjustment, identifier terminal and the readable storage medium storing program for executing of cog region |
| CN110751678A (en) * | 2018-12-12 | 2020-02-04 | 北京嘀嘀无限科技发展有限公司 | Moving object detection method and device and electronic equipment |
| CN109919008A (en) * | 2019-01-23 | 2019-06-21 | 平安科技（深圳）有限公司 | Moving target detecting method, device, computer equipment and storage medium |
| WO2020151172A1 (en) * | 2019-01-23 | 2020-07-30 | 平安科技（深圳）有限公司 | Moving object detection method and apparatus, computer device, and storage medium |
| CN112862855A (en) * | 2019-11-12 | 2021-05-28 | 北京京邦达贸易有限公司 | Image annotation method and device, computing equipment and storage medium |
| CN112862855B (en) * | 2019-11-12 | 2024-05-24 | 北京京邦达贸易有限公司 | Image labeling method, device, computing equipment and storage medium |
| CN113453055A (en) * | 2020-03-25 | 2021-09-28 | 华为技术有限公司 | Method and device for generating video thumbnail and electronic equipment |
| CN111626106A (en) * | 2020-04-17 | 2020-09-04 | 惠州市德赛西威智能交通技术研究院有限公司 | Camera vehicle detection rate statistical method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106991668B (en) | 2020-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106991668A (en) | A kind of evaluation method of day net camera shooting picture | |
| CN110136447B (en) | Method for detecting lane change of driving and identifying illegal lane change | |
| CN105744232B (en) | A kind of method of the transmission line of electricity video external force damage prevention of Behavior-based control analytical technology | |
| CN104966045B (en) | Aircraft disengaging berth automatic testing method based on video | |
| CN104378582B (en) | A kind of intelligent video analysis system and method cruised based on Pan/Tilt/Zoom camera | |
| CN109431523B (en) | Autism primary screening device based on non-social voice stimulation behavior paradigm | |
| CN105574543B (en) | A kind of vehicle brand type identifier method and system based on deep learning | |
| CN105913685A (en) | Video surveillance-based carport recognition and intelligent guide method | |
| CN103632158B (en) | Forest fire prevention monitor method and forest fire prevention monitor system | |
| Rachmadi et al. | Adaptive traffic signal control system using camera sensor and embedded system | |
| CN105208325B (en) | The land resources monitoring and early warning method captured and compare analysis is pinpointed based on image | |
| CN107229929A (en) | A kind of license plate locating method based on R CNN | |
| CN105469105A (en) | Cigarette smoke detection method based on video monitoring | |
| CN100454355C (en) | Video method for collecting information of vehicle flowrate on road in real time | |
| DE102015200589A1 (en) | Improved video-based system for automated detection of second-in-class parking violations | |
| CN107194396A (en) | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system | |
| CN103425967A (en) | Pedestrian flow monitoring method based on pedestrian detection and tracking | |
| CN109145736B (en) | A kind of detection method that the subway station pedestrian based on video analysis inversely walks | |
| CN106339687A (en) | Self-adaptive pedestrian street-crossing signal calculating method based on video | |
| CN109214345A (en) | The method for searching the driving trace for changing board vehicle is compared based on similarity | |
| CN110659546B (en) | Illegal booth detection method and device | |
| DE102006053286A1 (en) | Method for detecting movement-sensitive image areas, apparatus and computer program for carrying out the method | |
| CN106570440A (en) | People counting method and people counting device based on image analysis | |
| CN113450573A (en) | Traffic monitoring method and traffic monitoring system based on unmanned aerial vehicle image recognition | |
| CN111738336A (en) | Image detection method based on multi-scale feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |