CN109785386A - Object identification localization method and device - Google Patents
Object identification localization method and device Download PDFInfo
- Publication number
- CN109785386A CN109785386A CN201711118564.5A CN201711118564A CN109785386A CN 109785386 A CN109785386 A CN 109785386A CN 201711118564 A CN201711118564 A CN 201711118564A CN 109785386 A CN109785386 A CN 109785386A
- Authority
- CN
- China
- Prior art keywords
- monitor video
- video image
- foreground mask
- existence
- image
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The present invention provides a kind of object identification localization method and devices, method therein includes: the background image for obtaining monitor video image, the foreground mask figure of monitor video image is obtained based on background image, the connected region in foreground mask figure is obtained, based on the foreground mask figure and monitor video image recognition in connected region and positions object.Object identification localization method and device of the invention, foreground mask is capable of providing more information for categorised decision together with original picture block input classifier, extracting foreground object position using background modeling technology can be improved the speed of object positioning, and the accuracy and efficiency of identification positioning can be improved.
Description
Technical field
The present invention relates to technical field of image processing more particularly to a kind of object identification localization method and devices.
Background technique
Video monitoring is the important component of safety and protection system, including front-end camera, transmission cable, video monitoring
Platform.Video camera can be divided into network digital camera and analog video camera, can be used as the acquisition of head end video picture signal.Video
Monitor supervision platform carries out automatic identification, storage and automatic alarm etc. to image.User can watch image in real time, typing,
The operation such as play back, recall and store.Object positioning and identification are position and the type that object is oriented in image or video frame,
It is important video monitoring basic technology.In existing technology, object positioning needs picture carrying out different scale with identification
Scaling, scanning search is then carried out on the image of different scale or video frame, therefore calculation amount is very big, use cost
It is high.
Summary of the invention
In view of this, one or more embodiments of the invention provides a kind of object identification localization method and device.
According to one aspect of the disclosure, a kind of object identification localization method is provided, comprising: located in advance to monitor video
Reason obtains the background image of monitor video image；The foreground mask of the monitor video image is obtained based on the background image
Figure；The connected region in the foreground mask figure is obtained, the connected region is set as object domain of the existence；Based on the object
Foreground mask figure and the monitor video image recognition and positioning in domain of the existence are located in the object domain of the existence
Object.
Optionally, the foreground mask figure based in the object domain of the existence and the monitor video image recognition
And position be located at the object domain of the existence in object include: by the object domain of the existence foreground mask figure and
Local video image corresponding with the object domain of the existence is sent to classifier in the monitor video image, so that described point
Class device identifies and positions the object in the object domain of the existence.
Optionally, the classifier includes: ICF+AdaBoost classifier, DPM+LSVM classifier.
Optionally, the background image for obtaining monitor video image includes: to be solved using decoder to monitor video
Code obtains multiframe monitor video image；Based on the multiframe monitor video image and according to the foundation of preset model algorithm
Background model；The background image of the monitor video image is obtained according to the background model.
Optionally, the background model includes: gauss hybrid models；By the gauss hybrid models, that treated is described
The data of monitor video image include: RGB data, foreground mask data.
Optionally, the foreground mask figure for obtaining the monitor video image based on the background image includes: by institute
It states monitor video image and subtracts the background image, obtain the foreground mask figure.
Optionally, the connected region obtained in the foreground mask figure includes: based on preset algorithm to the institute
It states foreground mask figure to be handled, obtains the connected region obtained in the foreground mask figure；Wherein, the preset algorithm packet
It includes: erosion algorithm.
According to another aspect of the present disclosure, a kind of object identification positioning device is provided, comprising: background obtains module, is used for
Monitor video is pre-processed, the background image of monitor video image is obtained；Prospect obtains module, for being based on the background
Image obtains the foreground mask figure of the monitor video image；Connected region obtains module, for obtaining the foreground mask figure
In connected region, the connected region is set as object domain of the existence；Recognition processing module, for being existed based on the object
Foreground mask figure and the monitor video image recognition and positioning in region are located at the object in the object domain of the existence.
Optionally, the recognition processing module, for by foreground mask figure in the object domain of the existence and described
Local video image corresponding with the object domain of the existence is sent to classifier in monitor video image, so that the classifier
It identifies and positions the object in the object domain of the existence.
Optionally, the classifier includes: ICF+AdaBoost classifier, DPM+LSVM classifier.
Optionally, the background obtains module, for being decoded using decoder to monitor video, obtains multiframe monitoring
Video image；Based on the multiframe monitor video image and the background model is established according to preset model algorithm；According to institute
State the background image that background model obtains the monitor video image.
Optionally, the background model includes: gauss hybrid models；By the gauss hybrid models, that treated is described
The data of monitor video image include: RGB data, foreground mask data.
Optionally, the prospect obtains module, for the monitor video image to be subtracted the background image, obtains institute
State foreground mask figure.
Optionally, the connected region obtains module, for based on preset algorithm to the foreground mask figure into
Row processing, acquisition obtain the connected region in the foreground mask figure wherein, and the preset algorithm includes: erosion algorithm.
According to the another aspect of the disclosure, a kind of object identification positioning device is provided, comprising: memory；And it is coupled to
The processor of the memory, the processor is configured to the instruction based on storage in the memory, executes institute as above
The object identification localization method stated.
According to the another further aspect of the disclosure, a kind of computer readable storage medium is provided, computer program is stored thereon with
The step of instruction, which realizes method as described above when being executed by one or more processors.
The object identification localization method and device of the disclosure, object identification localization method and device obtain monitor video figure
The background image of picture obtains the foreground mask figure of monitor video image based on background image, obtains the connection in foreground mask figure
Region based on the foreground mask figure and monitor video image recognition in connected region and positions object；By foreground mask together with
Original picture block input classifier is capable of providing more information for categorised decision, extracts foreground object using background modeling technology
Position can be improved the speed of object positioning, and the accuracy and efficiency of identification positioning can be improved.
Detailed description of the invention
In order to illustrate more clearly of the embodiment of the present disclosure or technical solution in the prior art, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only
Some embodiments of the present disclosure, for those of ordinary skill in the art, without any creative labor, also
Other drawings may be obtained according to these drawings without any creative labor.
Fig. 1 is the flow diagram according to one embodiment of the object identification localization method of the disclosure；
Fig. 2A is to be illustrated to be intended to according to the foreground mask in one embodiment of the object identification localization method of the disclosure；Figure
2B is according to the positioning in one embodiment of the object identification localization method of the disclosure and schematic diagram of classifying；
Fig. 3 is the module diagram according to one embodiment of the object identification positioning device of the disclosure；
Fig. 4 is the module diagram according to another embodiment of the object identification positioning device of the disclosure.
Specific embodiment
The disclosure is described more fully with reference to the accompanying drawings, wherein illustrating the exemplary embodiment of the disclosure.Under
Face will combine the attached drawing in the embodiment of the present disclosure, and the technical solution in the embodiment of the present disclosure is clearly and completely described, and show
So, described embodiment is only disclosure a part of the embodiment, instead of all the embodiments.Based on the reality in the disclosure
Example is applied, every other embodiment obtained by those of ordinary skill in the art without making creative efforts all belongs to
In the range of disclosure protection.
Fig. 1 is according to the flow diagram of one embodiment of the object identification localization method of the disclosure, as shown in Figure 1:
Step 101, monitor video is pre-processed, obtains the background image of monitor video image.
Step 102, the foreground mask figure of monitor video image is obtained based on background image.
Step 103, the connected region in foreground mask figure is obtained, connected region is set as object domain of the existence.
Step 104, based on the foreground mask figure and monitor video image recognition in object domain of the existence and it is located at
Object in object domain of the existence.Object includes pedestrian, automobile etc..
In one embodiment, the background image for obtaining monitor video image can be there are many method.For example, using decoding
Device is decoded monitor video, obtains multiframe monitor video image, based on multiframe monitor video image and according to preset mould
Type algorithm establishes background model, and the background image of monitor video image is obtained according to background model.
Background model can there are many, for example, gauss hybrid models etc..Gauss hybrid models utilize Gaussian probability density
Function (normal distribution curve) accurately quantifies things, things is decomposed into several based on Gaussian probability-density function (normal state
Distribution curve) formed model.Gauss hybrid models Gaussian probability-density function accurately quantifies things, by a things point
Solution is several basic Gaussian probability-density function.Data by gauss hybrid models treated monitor video image include
RGB data, foreground mask data etc..
The camera that monitoring class video uses generally is fixedly installed, and therefore, the background of monitor video is within a short period of time
It is also relatively-stationary, and background image can be obtained by technological means from a series of monitored pictures.For example, to multiple
Monitoring image carries out average computation, and foreground object is handled by average computation and achievees the effect that abatement, or mixed using Gauss
Molding type is more accurately estimated background.Monitor video is decoded by decoder first, and video is reduced to a frame frame
Image, decoded picture frame is inputed into gauss hybrid models, by gauss hybrid models obtain current monitor image back
Scape image.
The foreground mask figure for obtaining monitor video image can be there are many method.For example, monitor video image is subtracted back
Scape image obtains foreground mask figure.After background image acquisition, " current image frame-can be passed through to current monitor picture
Background image " obtains foreground mask figure, as shown in Figure 2 A.The background of currently processed picture frame is obtained using gauss hybrid models,
Currently processed picture frame is subtracted into background image and obtains foreground image.Foreground mask figure it is obvious prompted foreground object
Existence, therefore, the region for object positioning provide accurate reference.
Foreground mask figure is handled based on preset algorithm, obtains the connected region obtained in foreground mask figure.In advance
If algorithm can there are many, for example, erosion algorithm etc..Foreground mask figure connected region can be obtained by erosion algorithm, it will
Each connected region is as possible object space.Binaryzation can also be carried out to foreground mask figure, and utilizes Mathematical Morphology side
Method carries out Denoising disposal to foreground image.Mathematical morphology is to go measurement and extraction figure with the structural element with certain form
Correspondingly-shaped as in is to achieve the purpose that image analysis and identification.
In one embodiment, by the foreground mask figure and monitor video image in object domain of the existence with object
The corresponding local video image of domain of the existence is sent to classifier, so that classifier is identified and positioned in object domain of the existence
Object, including pedestrian, automobile etc..Classifier includes ICF+AdaBoost classifier, DPM+LSVM classifier etc..Adaboost
It is a kind of iterative algorithm, for the different classifier (Weak Classifier) of same training set training, then these Weak Classifiers
It gathers, constitutes a stronger final classification device (strong classifier).ICF (Integral Channel Features, product
Subchannel feature) it is to take its rectangle frame at random in histogram of gradients by the way of Haar feature on the basis of HOG feature
Feature, and joined the integrating channel feature of the channel L and gradient channel.DPM (Deformable Parts Model, changeability
Partial model) utilize pyramid to extract HOG feature on different resolution.
By monitor video image and speculate that the background image obtained carries out residual noise reduction, obtains foreground image, present scene
After connected region where body is positioned to, the image by connected region mask together with corresponding monitored picture position is sent jointly to
Classifier differentiates type for classifier, as shown in Figure 2 B.For colored monitor video, it is sent to the image data shape of classifier
It can be classification since mask provides certain profile information at four-way information (RGB channel+mask channel)
Device provides more inputs, to promote classification accuracy.
As shown in figure 3, the disclosure provides a kind of object identification positioning device 30, comprising: background obtains module 31, prospect obtains
Modulus block 32, connected region obtain module 33 and recognition processing module 34.
Background obtains module 31 and pre-processes to monitor video, obtains the background image of monitor video image.Prospect obtains
Modulus block 32 obtains the foreground mask figure of monitor video image based on background image.Connected region obtains 33 acquisition prospect of module and covers
Connected region in code figure, is set as object domain of the existence for connected region.Recognition processing module 34 is based in object domain of the existence
Foreground mask figure and monitor video image recognition and position be located at object domain of the existence in object.
In one embodiment, background is obtained module 31 and is decoded using decoder to monitor video, obtains multiframe prison
Video image is controlled, establishes background model based on multiframe monitor video image and according to preset model algorithm.Background obtains module
31 obtain the background image of monitor video image according to background model.Background model includes gauss hybrid models etc., by Gauss
The data of mixed model treated monitor video image include RGB data, foreground mask data etc..
Prospect obtains module 32 and monitor video image is subtracted background image, obtains foreground mask figure.Connected region obtains
Module 33 is based on preset algorithm and handles foreground mask figure, and acquisition obtains the connected region in foreground mask figure wherein,
Preset algorithm includes erosion algorithm etc..
Recognition processing module 34 will be deposited in the foreground mask figure and monitor video image in object domain of the existence with object
In region, corresponding local video image is sent to classifier, so that classifier is identified and positioned in object domain of the existence
Object.Classifier includes: ICF+AdaBoost classifier, DPM+LSVM classifier etc..
Fig. 4 is the module diagram according to another embodiment of network side equipment disclosed by the invention.As shown in figure 4,
The device may include memory 41, processor 42, communication interface 43 and bus 44.Memory 41 for storing instruction, is handled
Device 42 is coupled to memory 41, and processor 42 is configured as realizing that above-mentioned object is known based on the instruction execution that memory 41 stores
Other localization method.
Memory 41 can be high speed RAM memory, nonvolatile memory (NoN-volatile memory) etc., deposit
Reservoir 41 is also possible to memory array.Memory 41 is also possible to by piecemeal, and block can be combined into virtually by certain rule
Volume.Processor 42 can be central processor CPU or application-specific integrated circuit ASIC (Application Specific
Integrated Circuit), or it is arranged to implement one or more of object identification localization method disclosed by the invention
A integrated circuit.
In one embodiment, the disclosure also provides a kind of computer readable storage medium, wherein computer-readable storage
Media storage has computer instruction, and the object identification positioning side that any embodiment as above is related to is realized in instruction when being executed by processor
Method.It should be understood by those skilled in the art that, embodiment of the disclosure can provide as method, apparatus or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the disclosure
Form.Moreover, can be used can be with non-in the computer that one or more wherein includes computer usable program code for the disclosure
The computer program implemented on instantaneity storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The disclosure is reference according to the method for the embodiment of the present disclosure, the flow chart of equipment (system) and computer program product
And/or block diagram describes.It should be understood that each process in flowchart and/or the block diagram can be realized by computer program instructions
And/or the combination of the process and/or box in box and flowchart and/or the block diagram.It can provide these computer programs to refer to
Enable the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to generate
One machine so that by the instruction that the processor of computer or other programmable data processing devices executes generate for realizing
The device for the function of being specified in one or more flows of the flowchart and/or one or more blocks of the block diagram.
So far, the disclosure is described in detail.In order to avoid covering the design of the disclosure, it is public that this field institute is not described
The some details known.Those skilled in the art as described above, completely it can be appreciated how implementing technology disclosed herein
Scheme.
Object identification localization method and device provided by the above embodiment obtains the background image of monitor video image, base
The foreground mask figure of monitor video image is obtained in background image, obtains the connected region in foreground mask figure, is based on connected region
Foreground mask figure and monitor video image recognition and positioning in domain are located at the object in object domain of the existence；By foreground mask
More information are capable of providing for categorised decision together with original picture block input classifier, extract prospect using background modeling technology
Object space can be improved the speed of object positioning, and the accuracy and efficiency of identification positioning, and easy to operate, energy can be improved
Enough reduce cost.
Disclosed method and system may be achieved in many ways.For example, can by software, hardware, firmware or
Software, hardware, firmware any combination realize disclosed method and system.The said sequence of the step of for method is only
In order to be illustrated, the step of disclosed method, is not limited to sequence described in detail above, especially says unless otherwise
It is bright.In addition, in some embodiments, also the disclosure can be embodied as to record program in the recording medium, these programs include
For realizing according to the machine readable instructions of disclosed method.Thus, the disclosure also covers storage for executing according to this public affairs
The recording medium of the program for the method opened.
The description of the disclosure is given for the purpose of illustration and description, and is not exhaustively or by the disclosure
It is limited to disclosed form.Many modifications and variations are obvious for the ordinary skill in the art.It selects and retouches
Embodiment is stated and be the principle and practical application in order to more preferably illustrate the disclosure, and those skilled in the art is enable to manage
The solution disclosure is to design various embodiments suitable for specific applications with various modifications.
Claims (16)
1. a kind of object identification localization method, comprising:
Monitor video is pre-processed, the background image of monitor video image is obtained；
The foreground mask figure of the monitor video image is obtained based on the background image；
The connected region in the foreground mask figure is obtained, the connected region is set as object domain of the existence；
Based on the foreground mask figure and the monitor video image recognition in the object domain of the existence and position positioned at described
Object in object domain of the existence.
2. the method for claim 1, wherein foreground mask figure and institute based in the object domain of the existence
It states monitor video image recognition and positions the object being located in the object domain of the existence and include:
By in the foreground mask figure and the monitor video image in the object domain of the existence with the object domain of the existence
Corresponding local video image is sent to classifier, so that the classifier is identified and positioned in the object domain of the existence
Object.
3. method according to claim 2, wherein
The classifier includes: ICF+AdaBoost classifier, DPM+LSVM classifier.
4. the method for claim 1, wherein the background image for obtaining monitor video image includes:
Monitor video is decoded using decoder, obtains multiframe monitor video image；
Based on the multiframe monitor video image and the background model is established according to preset model algorithm；
The background image of the monitor video image is obtained according to the background model.
5. method as claimed in claim 4, wherein the background model includes: gauss hybrid models；
Data by the gauss hybrid models treated the monitor video image include: RGB data, foreground mask number
According to.
6. method as claimed in claim 5, wherein it is described the monitor video image is obtained based on the background image before
Scape mask figure includes:
The monitor video image is subtracted into the background image, obtains the foreground mask figure.
7. the method for claim 1, wherein the connected region obtained in the foreground mask figure includes:
The foreground mask figure is handled based on preset algorithm, obtains the connection obtained in the foreground mask figure
Region
Wherein, the preset algorithm includes: erosion algorithm.
8. a kind of object identification positioning device, wherein include:
Background obtains module, for pre-processing to monitor video, obtains the background image of monitor video image；
Prospect obtains module, for obtaining the foreground mask figure of the monitor video image based on the background image；
Connected region obtains module and the connected region is set as object for obtaining the connected region in the foreground mask figure
Body domain of the existence；
Recognition processing module, for based in the object domain of the existence foreground mask figure and the monitor video image know
Not and position the object being located in the object domain of the existence.
9. device as claimed in claim 8, wherein
The recognition processing module, for by the foreground mask figure and the monitor video figure in the object domain of the existence
Local video image corresponding with the object domain of the existence is sent to classifier as in, so that the classifier is identified and positioned
Object in the object domain of the existence.
10. device as claimed in claim 9, wherein
The classifier includes: ICF+AdaBoost classifier, DPM+LSVM classifier.
11. device as claimed in claim 8, wherein
The background obtains module, for being decoded using decoder to monitor video, obtains multiframe monitor video image；Base
In the multiframe monitor video image and the background model is established according to preset model algorithm；It is obtained according to the background model
Obtain the background image of the monitor video image.
12. device as claimed in claim 11, wherein the background model includes: gauss hybrid models；By the Gauss
The data of mixed model treated the monitor video image include: RGB data, foreground mask data.
13. device as claimed in claim 12, wherein
The prospect obtains module and obtains the foreground mask for the monitor video image to be subtracted the background image
Figure.
14. device as claimed in claim 8, wherein
The connected region obtains module, for being handled based on preset algorithm the foreground mask figure, obtains
Obtain connected region in the foreground mask figure wherein, the preset algorithm includes: erosion algorithm.
15. a kind of object identification positioning device, comprising:
Memory；And it is coupled to the processor of the memory, the processor is configured to based on the storage is stored in
Instruction in device executes the object identification localization method as described in any one of claims 1 to 7.
16. a kind of computer readable storage medium, is stored thereon with computer program instructions, which is handled by one or more
The step of method described in claim 1 to 7 any one is realized when device executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711118564.5A CN109785386A (en) | 2017-11-14 | 2017-11-14 | Object identification localization method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711118564.5A CN109785386A (en) | 2017-11-14 | 2017-11-14 | Object identification localization method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109785386A true CN109785386A (en) | 2019-05-21 |
Family
ID=66493970
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711118564.5A Pending CN109785386A (en) | 2017-11-14 | 2017-11-14 | Object identification localization method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109785386A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110706251A (en) * | 2019-09-03 | 2020-01-17 | 北京正安维视科技股份有限公司 | Cross-lens tracking method for pedestrians |
| CN111131812A (en) * | 2019-12-31 | 2020-05-08 | 北京奇艺世纪科技有限公司 | Broadcast time testing method and device and computer readable storage medium |
| CN113393490A (en) * | 2020-03-12 | 2021-09-14 | 中国电信股份有限公司 | Target detection method and device, and computer-readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106557760A (en) * | 2016-11-28 | 2017-04-05 | 江苏鸿信***集成有限公司 | Monitoring system is filtered in a kind of image frame retrieval based on video identification technology |
| CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
-
2017
- 2017-11-14 CN CN201711118564.5A patent/CN109785386A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106557760A (en) * | 2016-11-28 | 2017-04-05 | 江苏鸿信***集成有限公司 | Monitoring system is filtered in a kind of image frame retrieval based on video identification technology |
| CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110706251A (en) * | 2019-09-03 | 2020-01-17 | 北京正安维视科技股份有限公司 | Cross-lens tracking method for pedestrians |
| CN110706251B (en) * | 2019-09-03 | 2022-09-23 | 北京正安维视科技股份有限公司 | Cross-lens tracking method for pedestrians |
| CN111131812A (en) * | 2019-12-31 | 2020-05-08 | 北京奇艺世纪科技有限公司 | Broadcast time testing method and device and computer readable storage medium |
| CN113393490A (en) * | 2020-03-12 | 2021-09-14 | 中国电信股份有限公司 | Target detection method and device, and computer-readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108921782B (en) | Image processing method, device and storage medium | |
| CN111370020B (en) | Method, system, device and storage medium for converting voice into lip shape | |
| CN109272509B (en) | Target detection method, device and equipment for continuous images and storage medium | |
| CN108154105B (en) | Underwater biological detection and identification method and device, server and terminal equipment | |
| CN105095905B (en) | Target identification method and Target Identification Unit | |
| CN110070029B (en) | Gait recognition method and device | |
| CN112597864B (en) | Monitoring video anomaly detection method and device | |
| CN112183482A (en) | Dangerous driving behavior recognition method, device and system and readable storage medium | |
| CN109740589B (en) | Asynchronous object ROI detection method and system in video mode | |
| CN112052815B (en) | Behavior detection method and device and electronic equipment | |
| CN109785386A (en) | Object identification localization method and device | |
| CN111067522A (en) | Brain addiction structural map assessment method and device | |
| CN113780243B (en) | Training method, device, equipment and storage medium for pedestrian image recognition model | |
| CN109815936A (en) | A kind of target object analysis method and device, computer equipment and storage medium | |
| CN103108124A (en) | Image acquisition method, device and mobile terminal | |
| CN111325773A (en) | Method, device and equipment for detecting moving target and readable storage medium | |
| CN111723656B (en) | Smog detection method and device based on YOLO v3 and self-optimization | |
| CN105844204B (en) | Human behavior recognition method and device | |
| CN111414886A (en) | Intelligent recognition system for human body dynamic characteristics | |
| EP4332910A1 (en) | Behavior detection method, electronic device, and computer readable storage medium | |
| CN111222370A (en) | Case studying and judging method, system and device | |
| CN113313124B (en) | Method and device for identifying license plate number based on image segmentation algorithm and terminal equipment | |
| CN115424253A (en) | License plate recognition method and device, electronic equipment and storage medium | |
| CN114004974A (en) | Method and device for optimizing images shot in low-light environment | |
| CN114612907A (en) | License plate recognition method and device |
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: 20190521 |