CN106846368A - A kind of oil field real-time video Intelligent Measurement and tracking and device - Google Patents
A kind of oil field real-time video Intelligent Measurement and tracking and device Download PDFInfo
- Publication number
- CN106846368A CN106846368A CN201710046060.0A CN201710046060A CN106846368A CN 106846368 A CN106846368 A CN 106846368A CN 201710046060 A CN201710046060 A CN 201710046060A CN 106846368 A CN106846368 A CN 106846368A
- Authority
- CN
- China
- Prior art keywords
- target
- frame
- tracking
- real
- template
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Mining
-
- 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
Abstract
The present invention relates to a kind of oil field real-time video Intelligent Measurement and tracking and device.The method includes:Obtain video flowing;Target detection is carried out to video flowing;When there is target in object detection results, real-time tracking is carried out to target.Be combined for target detection, real-time tracking, oil field business by the present invention, moving target is extracted quickly, accurately, while algorithm is simple, it is easy to implement, the trend of target can be understood with real-time capture, the movement position of positioning target, ensure In Oil Field Exploration And Development production safety.
Description
Technical field
The present invention relates to a kind of detection of video intelligent and tracking, intelligently examined more particularly, to a kind of oil field real-time video
Survey and tracking and device.
Background technology
The facilities such as the well site in oil field, storehouse of standing are threatened by some factors leading to social instability sometimes, for example row human or animal mesh
Target is swarmed into.In order to ensure exploration, exploitation, the production safety in oil field, it is necessary to which the real-time video to passing back carries out intelligence point
Analysis.If it find that there is swarming into for Unknown Subject, detect in time, so as to follow-up treatment.
Existing various video detecting methods can be to the various places required for people, there is provided a kind of real-time, vivid, true
The picture of the monitored object of real reflection, it is various that these video monitoring systems are integrated with prevention, monitoring, control evidence obtaining and management etc.
Function, can be used as treatment or ex-post analysis immediately.
However, existing video detection is not directed to oil field business with tracking.For example for the detection of vehicle
With tracking:Vehicle in detection video, to the car number for marking, and recording information of vehicles, if driver requested right
Vehicle is tracked, then be switched to tracing mode.But this method is primarily directed to the detection and tracking of vehicle.Such as pin again
Detection and tracking to water conservancy prevention and control, " the Flood Prevention monitoring based on video monitoring is pre- as disclosed in CN103325216A
Intelligent high-speed ball-shaped camera is connected to video behavior Analysis server, User logs in prison by alarm method and system ", the method
In the middle of control early warning platform software, early warning image, the store path of information are set with storage in the memory cell of monitoring and early warning platform
Mode, video behavior Analysis server gradually calls the intelligent high-speed ball machine presetting bit of association, and it is related to video monitoring and river
The combination of flood control blowdown detection.But this method is only for water conservancy prevention and control, at present not oilfield carry out attempt with
Explore.
For oil field, real-time monitoring is carried out primarily directed to the oil well, oil pipeline and station storehouse being distributed in the wild,
Any movable body may all be damaged to these facilities, and from for this angle, oil field emergency monitoring can more pay close attention to motion mesh
Target is detected and its follow-up trend.
The content of the invention
(1) technical problem to be solved
In order to solve the deficiencies in the prior art, the present invention provide a kind of oil field real-time video Intelligent Measurement and tracking and
Device, target detection, real-time tracking, oil field business are combined, and extract moving target quickly, accurately, while algorithm is simple, it is real
Apply conveniently, the trend of target can be understood with real-time capture, the movement position of positioning target, ensure In Oil Field Exploration And Development production peace
Entirely.
(2) technical scheme
In order to achieve the above object, the main technical schemes that the present invention is used include:
A kind of oil field real-time video Intelligent Measurement and tracking, it includes:
101, obtain video flowing;
102, target detection is carried out to the video flowing;
103, when there is target in object detection results, real-time tracking is carried out to the target.
Alternatively, step 102, specifically includes:
102-1, for the video flowing in any one frame, obtain any one frame previous frame image and described
Anticipate next two field picture of a frame;
102-2, calculates the poor D (n, n-1) and the next frame of any one frame and the previous frame with described
Anticipate a frame poor D (n+1, n);
102-3, according to the D (n, n-1), the D, (n+1, relation n) and between predetermined threshold value A extracts moving target
Binary image D (n).
Alternatively, step 102-2, specifically includes:
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture.
Alternatively, step 102-3, specifically includes:
Alternatively, step 103, specifically includes:
103-1, the size according to any one two field picture S sets up template;
103-2, each pixel (i, j) is begun stepping through from the S upper left corners, is calculated the template and is covered at (i, j)
Image-region S (x, y) normalization correlation NC.
Alternatively, step 103-1, specifically includes:The size of the size=0.39S of the template.
Alternatively, step 103-2, specifically includes:
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is the S at (x+i, y+j) place
Brightness value.
Alternatively, after step 103 is performed, also include:
If not finding moving target after carrying out real-time tracking to the target, step 101 and subsequent step are re-executed.
In addition, the main technical schemes that the present invention is used also include:
A kind of oil field real-time video Intelligent Measurement and tracks of device, described device, including:
Acquisition module, for obtaining video flowing;
Detection module, for carrying out target detection to the video flowing that the acquisition module is obtained;
Tracking module, for when there is target in the object detection results of the detection module, being carried out to the target
Real-time tracking;
The detection module, for any one frame in for the video flowing, obtains the previous frame of any one frame
Next two field picture of image and any one frame;Any one frame is calculated with the poor D (n, n-1) of the previous frame and institute
State next frame and any one frame poor D (n+1, n);According to the D (n, n-1), the D (n+1, n) with predetermined threshold value A it
Between relation extract moving target binary image D (n);
The detection module, for according to equation below calculate D (n, n-1) and D (n+1, n),
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture;
The detection module, for calculating D (n) according to equation below,
The tracking module, for setting up template according to the size of any one two field picture S;Opened from the S upper left corners
The each pixel (i, j) of the traversal that begins, the normalization for calculating image-region S (x, y) that the template is covered at (i, j) is related
NC;The size of the size=0.39S of the template;
The tracking module, for calculating NC according to equation below,
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is the S at (x+i, y+j) place
Brightness value.
Alternatively, described device, also includes:
Control module, for when moving target is not found after carrying out real-time tracking to the target, control acquisition module,
Detection module and tracking module are re-executed.
(3) beneficial effect
The beneficial effects of the invention are as follows:Video flowing is obtained, target detection is carried out to video flowing, when being deposited in object detection results
In target, real-time tracking is carried out to target, target detection, real-time tracking, oil field business are combined, extract moving target fast
It is fast, accurate, while algorithm is simple, it is easy to implement, the trend of target can be understood with real-time capture, the movement position of positioning target,
Ensure In Oil Field Exploration And Development production safety.
Brief description of the drawings
A kind of oil field real-time video Intelligent Measurement that Fig. 1 is provided for one embodiment of the invention and tracking flow chart;
Another oil field real-time video Intelligent Measurement and tracking flow that Fig. 2 is provided for one embodiment of the invention
Figure;
Fig. 3 illustrates for a kind of Region Matching track algorithm related based on normalization that one embodiment of the invention is provided
Figure;
Another oil field real-time video Intelligent Measurement and tracking flow that Fig. 4 is provided for one embodiment of the invention
Figure;
A kind of oil field real-time video Intelligent Measurement that Fig. 5 is provided for one embodiment of the invention and tracks of device structural representation
Figure.
Specific embodiment
In order to preferably explain the present invention, in order to understand, below in conjunction with the accompanying drawings, by specific embodiment, to this hair
It is bright to be described in detail.
Existing video detection is with tracking not for oil field business.In order to solve the problem, according to oil field
The characteristics of exploration and development is produced, the present invention provides a kind of oil field real-time video Intelligent Measurement and tracking and device, Ke Yitong
Cross acquisition video flowing, target detection carried out to video flowing, when there is target in object detection results, target is carried out in real time with
Track, target detection, real-time tracking, oil field business are combined, and extract moving target quickly, accurately, while algorithm is simple, are implemented
It is convenient, the trend of target can be understood with real-time capture, the movement position of positioning target, ensure In Oil Field Exploration And Development production safety.
A kind of oil field real-time video Intelligent Measurement and tracking provided referring to Fig. 1, the present embodiment, including:
101, obtain video flowing.
102, target detection is carried out to video flowing.
Alternatively, step 102, specifically includes:
102-1, for video flowing in any one frame, obtain the previous frame image and any one frame of any one frame
Next two field picture.
102-2, calculates any one frame with the poor D (n, n-1) and next frame of previous frame and the poor D (n of any one frame
+1,n)。
102-3, according to D (n, n-1), D, (n+1, relation n) and between predetermined threshold value A extracts the binaryzation of moving target
Image D (n).
Alternatively, step 102-2, specifically includes:
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture.
Alternatively, step 102-3, specifically includes:
103, when there is target in object detection results, real-time tracking is carried out to target.
Alternatively, step 103, specifically includes:
103-1, the size according to any one two field picture S sets up template.
103-2, each pixel (i, j), the image district that calculation template is covered at (i, j) are begun stepping through from the S upper left corners
The normalization correlation NC in domain S (x, y).
Alternatively, step 103-1, specifically includes:The size of the size=0.39S of template.
Alternatively, step 103-2, specifically includes:
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is brightness of the S at (x+i, y+j) place
Value.
Alternatively, after step 103 is performed, also include:
If not finding moving target after carrying out real-time tracking to target, step 101 and subsequent step are re-executed.
The beneficial effect of the present embodiment is:By obtaining video flowing, target detection is carried out to video flowing, when target detection knot
When there is target in fruit, real-time tracking is carried out to target, target detection, real-time tracking, oil field business are combined, extract motion
Target is quick, accurate, while algorithm is simple, it is easy to implement, target can be understood with real-time capture, the movement position of positioning target
Trend, ensure In Oil Field Exploration And Development production safety.
With reference to the flow shown in Fig. 2, the oil field real-time video Intelligent Measurement that the present invention is provided is entered with tracking
Row is illustrated again.
201, obtain video flowing.
Specifically, video flowing is stored among internal memory after reading video flowing.
202, target detection is carried out to video flowing.
Specifically, carrying out target detection using three-frame differencing.
The basic ideas of frame difference algorithm are to carrying out difference processing to adjacent two frame or multiple image in video sequence, so
It is compared to obtain the pixel of motion parts with the threshold value A being previously set afterwards.
When implementing, can be achieved by the steps of.
202-1, for video flowing in any one frame, obtain any one frame previous frame image and any one frame it is next
Two field picture.
202-2, calculate the poor D (n, n-1) and next frame of any one frame and previous frame and any one frame poor D (n+1,
n)。
Wherein, D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture.
202-3, according to D (n, n-1), D, (n+1, relation n) and between predetermined threshold value A extracts the binaryzation of moving target
Image D (n).
Wherein,
Three-frame differencing is adapted to the situation that background will not change in a short time, is detected when finding that target is moved
Come.
203, when there is target in object detection results, real-time tracking is carried out to target.
Specifically, carrying out real-time tracking with based on the related Region Matching track algorithm of normalization.
Based on the normalization that the related Region Matching track algorithm of normalization passes through calculation template and template institute overlay area
Coefficient correlation carries out matched jamming (referring to Fig. 3).
In the specific implementation, can be achieved by the steps of.
203-1, the size according to any one two field picture S sets up template.
The template of foundation is generally the small rectangles of length-width ratio S.
Alternatively, the size of the size=0.39S of template.
203-2, each pixel (i, j), the image district that calculation template is covered at (i, j) are begun stepping through from the S upper left corners
The normalization correlation NC in domain S (x, y).
Wherein,
T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is brightness values of the S at (x+i, y+j) place.
After pixel matching primitives one by one in template and S regions, a matrix on NC is obtained.NC matrix elements
Between zero and one, value is bigger, and expression matching effect is better for value, and the image-region of 0 expression template and covering does not have correlation, i.e.,
Matching effect is worst, and 1 represents the image-region correlation highest of template and covering, i.e. matching effect is best.Simultaneously by this
Normalization relevant treatment can reduce the influence for bringing of illuminance abrupt variation.
204, if not finding moving target after carrying out real-time tracking to target, re-execute step 201 and subsequent step.
If not finding moving target, 201 are gone to step, proceed live video stream reading.
Step 201 to the method described in step 204, by using three-frame differencing to based on the related region of normalization
The mode that Matching pursuitalgorithm is combined carries out oil field real-time video intelligent monitoring, for oil field, primarily directed to distribution
Oil well in the wild, oil pipeline and station storehouse carry out real-time monitoring, and any movable body may all be damaged to these facilities,
Detection and its follow-up trend of moving target can be more paid close attention to from oil field emergency monitoring for this angle, by various functions module
Combine, three-frame differencing is combined with oil field business, extract moving target quickly, accurately, while algorithm is simple, it is real
Apply conveniently.And track and mesh can be understood with real-time capture, the movement position of positioning target by the related Region Matching of normalization
Target trend, ensures In Oil Field Exploration And Development production safety.
Fig. 4 shows the oil field real-time video Intelligent Measurement of the present embodiment offer and the practical application flow of tracking.
After live video stream is accessed, video flowing is read by video image read module, the video of reading flows through moving object detection mould
Block carries out target detection, if it find that moving target, then be tracked by real-time tracking module, if not finding to move mesh
Mark, then carry out target detection, until this monitoring terminates again through moving object detection module.
The beneficial effect of the present embodiment is:By obtaining video flowing, target detection is carried out to video flowing, when target detection knot
When there is target in fruit, real-time tracking is carried out to target, target detection, real-time tracking, oil field business are combined, extract motion
Target is quick, accurate, while algorithm is simple, it is easy to implement, target can be understood with real-time capture, the movement position of positioning target
Trend, the artificial inspection number of times in scene such that it is able to reduce oil field critical facility, and can continuously running, find in real time and
Record breaks in target, compared with artificial inspection, can reduce the human resources input cost of safety guarantee work, and improve oil
The safety guarantee coefficient of field critical facility.
Based on same inventive concept, the present invention also provides a kind of oil field real-time video Intelligent Measurement and tracks of device, the dress
The principle for putting solve problem is similar to tracking to oil field real-time video Intelligent Measurement, therefore the implementation of the device may refer to
A kind of oil field real-time video Intelligent Measurement and the implementation of tracking, repeat part and repeat no more.
Referring to Fig. 5, the oil field real-time video Intelligent Measurement and tracks of device, including:
Acquisition module 501, for obtaining video flowing;
Detection module 502, for carrying out target detection to the video flowing that acquisition module 501 is obtained;
Tracking module 503, for when there is target in the object detection results of detection module 502, reality being carried out to target
When track;
Detection module 502, for any one frame in for video flowing, obtain any one frame previous frame image and should
Next two field picture of any one frame;Calculate the poor D (n, n-1) and next frame and any one frame of any one frame and previous frame
Poor D (n+1, n);According to D (n, n-1), D, (n+1, relation n) and between predetermined threshold value A extracts the binary picture of moving target
As D (n);
Detection module 502, for according to equation below calculate D (n, n-1) and D (n+1, n),
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture;
Detection module 502, for calculating D (n) according to equation below,
Tracking module 503, for setting up template according to the size of any one two field picture S;Begun stepping through from the S upper left corners each
Pixel (i, j), the normalization correlation NC of image-region S (x, y) that calculation template is covered at (i, j);The size of template=
The size of 0.39S;
Tracking module 503, for calculating NC according to equation below,
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is brightness of the S at (x+i, y+j) place
Value.
Alternatively, the device, also includes:
Control module, for when moving target is not found after carrying out real-time tracking to target, controlling acquisition module, detection
Module and tracking module are re-executed.
The beneficial effect of the present embodiment is:By obtaining video flowing, target detection is carried out to video flowing, when target detection knot
When there is target in fruit, real-time tracking is carried out to target, target detection, real-time tracking, oil field business are combined, extract motion
Target is quick, accurate, while algorithm is simple, it is easy to implement, target can be understood with real-time capture, the movement position of positioning target
Trend, the artificial inspection number of times in scene such that it is able to reduce oil field critical facility, and can continuously running, find in real time and
Record breaks in target, compared with artificial inspection, can reduce the human resources input cost of safety guarantee work, and improve oil
The safety guarantee coefficient of field critical facility.
Claims (10)
1. a kind of oil field real-time video Intelligent Measurement and tracking, it is characterised in that methods described, including:
101, obtain video flowing;
102, target detection is carried out to the video flowing;
103, when there is target in object detection results, real-time tracking is carried out to the target.
2. method according to claim 1, it is characterised in that step 102, specifically includes:
102-1, for the video flowing in any one frame, obtain the previous frame image of any one frame and described any one
Next two field picture of frame;
102-2, calculates any one frame any one with described with the poor D (n, n-1) of the previous frame and the next frame
Frame poor D (n+1, n);
102-3, according to the D (n, n-1), the D, (n+1, relation n) and between predetermined threshold value A extracts the two of moving target
Value image D (n).
3. method according to claim 2, it is characterised in that step 102-2, specifically includes:
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture.
4. method according to claim 2, it is characterised in that step 102-3, specifically includes:
5. method according to claim 1, it is characterised in that step 103, specifically includes:
103-1, the size according to any one two field picture S sets up template;
103-2, each pixel (i, j) is begun stepping through from the S upper left corners, calculates the figure that the template is covered at (i, j)
As the normalization correlation NC of region S (x, y).
6. method according to claim 5, it is characterised in that step 103-1, specifically includes:The size of the template=
The size of 0.39S.
7. method according to claim 5, it is characterised in that step 103-2, specifically includes:
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is the S in the bright of (x+i, y+j) place
Angle value.
8. method according to claim 1, it is characterised in that after step 103 is performed, also include:
If not finding moving target after carrying out real-time tracking to the target, step 101 and subsequent step are re-executed.
9. a kind of oil field real-time video Intelligent Measurement and tracks of device, it is characterised in that described device, including:
Acquisition module, for obtaining video flowing;
Detection module, for carrying out target detection to the video flowing that the acquisition module is obtained;
Tracking module, for when there is target in the object detection results of the detection module, being carried out in real time to the target
Tracking;
The detection module, for any one frame in for the video flowing, obtains the previous frame image of any one frame
With next two field picture of any one frame;Calculate the poor D (n, n-1) of any one frame and the previous frame and it is described under
One frame and any one frame poor D (n+1, n);According to the D (n, n-1), the D (n+1, n) and predetermined threshold value A between
Relation extracts binary image D (n) of moving target;
The detection module, for according to equation below calculate D (n, n-1) and D (n+1, n),
D (n, n-1)=| In(x,y)-In-1(x,y)|;
D (n+1, n)=| In+1(x,y)-In(x,y)|;
Wherein, (x, y) is the coordinate of pixel, In(x, y) is the pixel value of any one two field picture;
The detection module, for calculating D (n) according to equation below,
The tracking module, for setting up template according to the size of any one two field picture S;Since the S upper left corners time
Each pixel (i, j) is gone through, the normalization correlation NC of image-region S (x, y) that the template is covered at (i, j) is calculated;Institute
State the size of the size=0.39S of template;
The tracking module, for calculating NC according to equation below,
Wherein, T (i, j) is brightness value of the template at (i, j) place, and S (x+i, y+j) is the S in the bright of (x+i, y+j) place
Angle value.
10. device according to claim 9, it is characterised in that described device, also includes:
Control module, for when moving target is not found after carrying out real-time tracking to the target, controlling acquisition module, detection
Module and tracking module are re-executed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710046060.0A CN106846368A (en) | 2017-01-22 | 2017-01-22 | A kind of oil field real-time video Intelligent Measurement and tracking and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710046060.0A CN106846368A (en) | 2017-01-22 | 2017-01-22 | A kind of oil field real-time video Intelligent Measurement and tracking and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106846368A true CN106846368A (en) | 2017-06-13 |
Family
ID=59119324
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710046060.0A Pending CN106846368A (en) | 2017-01-22 | 2017-01-22 | A kind of oil field real-time video Intelligent Measurement and tracking and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106846368A (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160012608A1 (en) * | 2014-07-08 | 2016-01-14 | Nec Corporation | Object tracking device, object tracking method, and computer-readable medium |
CN105632115A (en) * | 2015-12-31 | 2016-06-01 | 中海石油(中国)有限公司天津分公司 | Offshore oilfield security system |
CN105654512A (en) * | 2015-12-29 | 2016-06-08 | 深圳羚羊微服机器人科技有限公司 | Target tracking method and device |
-
2017
- 2017-01-22 CN CN201710046060.0A patent/CN106846368A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160012608A1 (en) * | 2014-07-08 | 2016-01-14 | Nec Corporation | Object tracking device, object tracking method, and computer-readable medium |
CN105654512A (en) * | 2015-12-29 | 2016-06-08 | 深圳羚羊微服机器人科技有限公司 | Target tracking method and device |
CN105632115A (en) * | 2015-12-31 | 2016-06-01 | 中海石油(中国)有限公司天津分公司 | Offshore oilfield security system |
Non-Patent Citations (1)
Title |
---|
刘焱: "《油田卫星应急监测关键技术研究》", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107967451B (en) | Method for counting crowd of still image | |
Zeng et al. | Underwater target detection based on Faster R-CNN and adversarial occlusion network | |
Cui et al. | SDDNet: A fast and accurate network for surface defect detection | |
CN108492319B (en) | Moving target detection method based on deep full convolution neural network | |
CN106096577B (en) | A kind of target tracking method in camera distribution map | |
CN109886245A (en) | A kind of pedestrian detection recognition methods based on deep learning cascade neural network | |
CN107016357A (en) | A kind of video pedestrian detection method based on time-domain convolutional neural networks | |
CN103886325B (en) | Cyclic matrix video tracking method with partition | |
CN106204638A (en) | A kind of based on dimension self-adaption with the method for tracking target of taking photo by plane blocking process | |
CN113554125B (en) | Object detection apparatus, method and storage medium combining global and local features | |
CN108986138A (en) | Method for tracking target and equipment | |
CN105224912A (en) | Based on the video pedestrian detection and tracking method of movable information and Track association | |
CN107452015A (en) | A kind of Target Tracking System with re-detection mechanism | |
CN109376736A (en) | A kind of small video target detection method based on depth convolutional neural networks | |
Yang et al. | Counting crowds using a scale-distribution-aware network and adaptive human-shaped kernel | |
Zhao et al. | Image-based crowd stability analysis using improved multi-column convolutional neural network | |
Pan et al. | Key algorithms of video target detection and recognition in intelligent transportation systems | |
Dai et al. | Real-time safety helmet detection system based on improved SSD | |
Hu | Football player posture detection method combining foreground detection and neural networks | |
CN109102549A (en) | Detection method, device, computer equipment and the storage medium of image light source color | |
Gómez-Conde et al. | Simple human gesture detection and recognition using a feature vector and a real-time histogram based algorithm | |
He et al. | Real-time pedestrian warning system on highway using deep learning methods | |
CN106846368A (en) | A kind of oil field real-time video Intelligent Measurement and tracking and device | |
He et al. | High accuracy intelligent real-time framework for detecting infant drowning based on deep learning | |
CN110516559A (en) | Suitable for precisely monitor method for tracking target and device, computer equipment |
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: 20170613 |