CN109800685A - The determination method and device of object in a kind of video - Google Patents
The determination method and device of object in a kind of video Download PDFInfo
- Publication number
- CN109800685A CN109800685A CN201811648094.8A CN201811648094A CN109800685A CN 109800685 A CN109800685 A CN 109800685A CN 201811648094 A CN201811648094 A CN 201811648094A CN 109800685 A CN109800685 A CN 109800685A
- Authority
- CN
- China
- Prior art keywords
- image
- image information
- detection
- identification
- information
- 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 invention discloses a kind of determination method and devices of object in video.The described method includes: obtaining the video to be processed that monitoring device is shot within a preset period of time, if the first image is detection frame image, the type that object is respectively identified in the first image and the corresponding detection image information of each identification object are detected；If the first image is prediction frame image, according to the corresponding image information of each identification object in the second image, prediction each identification object corresponding forecast image information in the first image；According to detection image information and forecast image information of the same identification object in different frame image, the identification image of identification object is determined.Due to first judging whether detection image information and forecast image information are images to be recognized, the precision of the identification image of object to be identified is improved, the accuracy that the subsequent identification image using object to be identified is filed is further increased.
Description
Technical field
The present invention relates to a kind of determination method and devices of object in technical field of video processing more particularly to video.
Background technique
In today's society, for the needs of safeguard management, it is various public that monitoring device has been covered with street, community, building etc.
Occasion.When alert occurs, the image of suspect or suspected vehicles is first determined from the video data that monitoring device acquires, so
Suspect or suspected vehicles are tracked down and arrested according to the image of suspect or suspected vehicles by police afterwards.
In the prior art, monitoring device acquisition video flowing after, mostly use to every frame image in video flowing carry out detection and
Identification, determines the objects such as face or the vehicle in every frame image, then again matches the object detected in each frame image,
Determine all images an of object, this method is imitated due to needing to be detected and identified that calculation amount is larger to every frame image
Rate is low.
Summary of the invention
The embodiment of the present invention provides a kind of determination method and device of object in video, to solve to need in the prior art pair
The technical issues of every frame image detected and identified, calculation amount is larger, low efficiency.
The embodiment of the present invention provides a kind of determination method of object in video, which comprises
The video to be processed that monitoring device is shot within a preset period of time is obtained, the video to be processed includes N frame
Image；N is more than or equal to 2；
For the first image, if the first image is detection frame image, respectively identification pair in the first image is detected
The type of elephant and the corresponding detection image information of each identification object；If the first image is prediction frame image, according to second
The corresponding image information of each identification object in image predicts each identification object corresponding prediction in the first image
Image information；The first image is any of described N frame image, and second image is that the first image is adjacent
Image and have determined that or predict the corresponding image information of identification object；
According to detection image information and forecast image information of the same identification object in different frame image, the knowledge is determined
The identification image of other object.
Due to first judging whether detection image information and forecast image information are images to be recognized, then according to same wait know
The detection image information for having been determined as images to be recognized and forecast image information of other object, determine the identification figure of object to be identified
Picture further increases the subsequent identification image using object to be identified to improve the precision of the identification image of object to be identified
The accuracy filed.
In one possible implementation, if the first image is detection frame image, it is determined that each identification object pair
The detection image information answered, comprising:
Determine the first image be detection frame image when, to the first image carry out object detection, determine described in
Corresponding first image information of each first object in first image；
According to the corresponding third image information of each second object in third image, predict each second object described
Corresponding second image information in first image；The third image is the phase of the first image of monitoring device acquisition
Adjacent image and the third image are prediction frame image；
When the second image information of the second object and the first image information of the first object meet and impose a condition, institute is determined
It states the second object and first object is same identification object, and believed according to the first image information and second image
Breath determines the detection image information of the same identification object.
In one possible implementation, if the first image is detection frame image, it is determined that the first image
In respectively identify the type and the corresponding detection image information of each identification object of object, comprising:
If the first image is detection frame image, the first image is inputted into sorter model, determines described the
The type and the corresponding detection image information of each identification object of object are respectively identified in one image；Class of the classifier for differentiation
Type includes motor vehicle, non-motor vehicle and pedestrian.
In one possible implementation, according to same detection image information of the identification object in different frame image and
Forecast image information determines the identification image of the identification object, comprising:
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
Identification image of a image information as the identification object；Alternatively,
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
A image information, and the identification image for identifying object is generated according to the K image information.
The embodiment of the present invention provides a kind of determining device of object in video, and described device includes:
Acquiring unit, the video to be processed shot within a preset period of time for obtaining monitoring device are described wait locate
Managing video includes N frame image；N is more than or equal to 2；
Processing unit, if the first image is detection frame image, detects first figure for being directed to the first image
The type and the corresponding detection image information of each identification object of object are respectively identified as in；If the first image is prediction frame figure
Picture predicts each identification object in first figure then according to the corresponding image information of each identification object in the second image
The corresponding forecast image information as in；The first image is any of described N frame image, and second image is described
The adjacent image of first image and have determined that or predict the corresponding image information of identification object；
The processing unit is also used to detection image information and prediction according to same identification object in different frame image
Image information determines the identification image of the identification object.
In one possible implementation, the processing unit is specifically used for:
Determine the first image be detection frame image when, to the first image carry out object detection, determine described in
Corresponding first image information of each first object in first image；
According to the corresponding third image information of each second object in third image, predict each second object described
Corresponding second image information in first image；The third image is the phase of the first image of monitoring device acquisition
Adjacent image and the third image are prediction frame image；
When the second image information of the second object and the first image information of the first object meet and impose a condition, institute is determined
It states the second object and first object is same identification object, and believed according to the first image information and second image
Breath determines the detection image information of the same identification object.
In one possible implementation, the processing unit is specifically used for:
If the first image is detection frame image, the first image is inputted into sorter model, determines described the
The type and the corresponding detection image information of each identification object of object are respectively identified in one image；Class of the classifier for differentiation
Type includes motor vehicle, non-motor vehicle and pedestrian.
In one possible implementation, the processing unit is specifically used for:
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
Identification image of a image information as the identification object；Alternatively,
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
A image information, and the identification image for identifying object is generated according to the K image information.
The embodiment of the present application also provides a kind of device, which, which has, realizes that object is really in video as described above
Determine the function of method.The function can execute corresponding software realization by hardware, in a kind of possible design, the device packet
It includes: processor, transceiver, memory；The memory is for storing computer executed instructions, and the transceiver is for realizing the device
It is communicated with other communication entities, which is connect with the memory by the bus, when the apparatus is operative, the processing
Device executes the computer executed instructions of memory storage, so that the device executes in video as described above object really
Determine method.
The embodiment of the present invention also provides a kind of computer storage medium, stores software program in the storage medium, this is soft
Part program realizes view described in above-mentioned various possible implementations when being read and executed by one or more processors
The determination method of object in frequency.
The embodiment of the present invention also provides a kind of computer program product comprising instruction, when run on a computer,
So that computer executes the determination method of object in video described in above-mentioned various possible implementations.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly introduced.
Fig. 1 is the schematic diagram for the system architecture that the embodiment of the present invention is applicable in；
Fig. 2 is flow diagram corresponding to the determination method of object in a kind of video provided in an embodiment of the present invention；
Fig. 3 for the embodiment of the present invention provide it is a kind of prediction frame image in identify object determination method corresponding to process show
It is intended to；
Fig. 4 is that the embodiment of the invention provides a kind of structural schematic diagrams of the determining device of object in video.
Specific embodiment
The application is specifically described with reference to the accompanying drawings of the specification, the concrete operation method in embodiment of the method can also
To be applied in Installation practice.
The determination of object in the embodiment of the present invention can be applied to image filing, such as using monitoring device acquisition
When video flowing is filed, it can be determined in video flowing using the determination method of object in video in the embodiment of the present invention with a pair of
As the object box in every frame image, then using the corresponding object box of the object as the identification image of object, for subsequent
Filing.
Fig. 1 illustrates the schematic diagram of the applicable system architecture of the embodiment of the present invention, includes in the system architecture
Monitoring device 101, server 102.Monitoring device 101 acquires video flowing in real time, then by the video stream of acquisition to service
Device 102 includes the determining device of object in video in server 102, and server 102 obtains image from video flowing, then really
Determine the corresponding image-region of object in image.Monitoring device 101 is connect by wireless network with server 102, monitoring device
It is the electronic equipment, such as camera, video camera, video recorder etc. for having acquisition image function.Server 102 is a server
Or the server cluster or cloud computing center of several servers composition.
Based on system architecture shown in FIG. 1, Fig. 2 illustrates object in a kind of video provided in an embodiment of the present invention
Determination method corresponding to flow diagram, the process of this method can execute by the determining device of object in video, video
The determining device of middle object can be server 102 shown in FIG. 1, as shown in Fig. 2, specifically includes the following steps:
Step 201, the video to be processed that monitoring device is shot within a preset period of time is obtained.
Video to be processed may include N frame image；N is more than or equal to 2.
Further, accessed video to be processed can be monitoring device within a preset period of time in chronological order according to
The second image and the first image of secondary output, wherein the second image is exported earlier than the first image.
Step 202, for the first image, if the first image is detection frame image, respectively identification pair in the first image is detected
The type of elephant and the corresponding detection image information of each identification object.
The corresponding frame image of each image information, the first object to be identified be any frame image include it is any to be identified right
As.
Specifically, the image of video to be processed is marked in advance, the figure that will be detected in video to be processed
It is prediction frame image by the image tagged predicted as being labeled as detection frame image.Illustratively, one section of video is set
Include 10 frame images in stream, is detection frame image, the second frame image to the 4th frame by first frame image and the 5th frame image tagged
Image tagged is prediction frame image, and the 6th frame to the tenth frame flag is prediction frame image.
Since the image in the video flowing of monitoring device acquisition is first divided into detection frame image and non-detection frame image, therefore
When obtaining the first image, first judge whether the first image is detection frame image, if it is, to be identified in the first image of detection
Otherwise object is predicted the object to be identified in the first image using the object to be identified in the second image, therefore is not needed to every frame
Images to be recognized is detected and is identified, to reduce the calculation amount of the object to be identified in determining image, is improved simultaneously
Efficiency.
Further, if the first image is detection frame image, the first image can be inputted into sorter model, determines the
The type and the corresponding detection image information of each identification object of object are respectively identified in one image.Wherein, classifier is for differentiation
Type includes motor vehicle, non-motor vehicle and pedestrian.That is, respectively identification object type include motor vehicle, non-motor vehicle and
Pedestrian, further, it is each identify the corresponding detection image information of object include the corresponding detection image information of motor vehicle, it is non-maneuver
The corresponding detection image information of vehicle and the corresponding detection image information of pedestrian.
Specifically, it can detecte the type for respectively identifying object in the first image, determine each identification object in the first image
Corresponding detection image information.
It is possible to further first carry out object detection to the first image, determines and respectively identify that object is corresponding in the first image
Then the corresponding detection image region of each identification object is inputted sorter model, so that it is determined that respectively by detection image region respectively
Identify the corresponding type of object.Wherein, image-region can be the frames images with regular shape, or not have regular shape
Frames images.
Optionally, for any identification object, the key point of test object in the image-region of identification object.According to pass
Key point adjustment identification object corresponding image-region in the first image.
Specifically, the key point of object is the key point of object for identification, for example the key point of motor vehicle may include
License plate, vehicle window and wheel etc., the key point of pedestrian may include head, four limbs, the upper part of the body and lower part of the body etc..
Due to detecting the key point of object to be identified, it is then based on key point adjustment image-region, to make image-region
It is more acurrate.
Step 203, if the first image is prediction frame image, according to the corresponding image of each identification object in the second image
Information predicts each identification object corresponding forecast image information in the first image.
Wherein, the first image is any of N frame image, and the second image is that the first image of monitoring device acquisition is adjacent
Image and have determined that or predict the corresponding image information of identification object.Specifically, the second image can be detection frame figure
Picture or non-detection frame image.
It is possible to further elder generation according to the corresponding image-region of object is respectively identified in the second image, each identification object is predicted
The corresponding image-region in the first image, and then can predict each object to be identified in the first image in correspondence image region
Image information, i.e., each identification object corresponding image information in the first image.
Specifically, it for the corresponding image-region of any identification object in the second image, is determined from the first image
The similarity of image-region corresponding with the identification object is greater than the image-region of preset threshold, as the identification object first
Corresponding forecast image region in image.
For example, it sets in known second image and respectively identifies the corresponding frames images of object, for being identified in the second image
The corresponding frames images of object A are determined to be greater than preset threshold with the similarity of the frames images of object to be identified A from the first image
Frames images, as frames images of the object to be identified A in the first image.For example, setting object to be identified as motor vehicle, it is known that
The corresponding vehicle frame of motor vehicle in second image, for the corresponding vehicle frame of motor vehicle each in the second image, from the first image
In determine with the similarity of the vehicle frame be greater than preset threshold vehicle frame, as vehicle of the motor vehicle in the first image
Frame.
In a kind of possible embodiment, the apparent region of feature is selected from identification object, by the region and first
Image is compared, and the image-region for being greater than preset threshold with the similarity in the region is determined from the first image, then will
Image-region is used as identification object corresponding image-region in the first image after expanding.
For example, object to be identified is set as motor vehicle, it is known that the corresponding vehicle frame of motor vehicle in the second image, for
Vehicle frame in second image, the first apparent region of feature in selection vehicle frame, such as wheel and vehicle window, will be in vehicle frame
The apparent region of feature is compared with the first image, determines to be greater than default threshold with the similarity in the region from the first image
The image-region of value is used as the motor vehicle corresponding vehicle frame in the first image after then expanding image-region.
In view of the possible more than one of image-region that similarity is greater than preset threshold, the embodiment of the present invention can be further
The quantity for being greater than the image-region of preset threshold to similarity judges.
If similarity be greater than preset threshold image-region only one, by the similarity be greater than preset threshold image
Region is as identification object corresponding image-region in the first image.
If the image-region that similarity is greater than preset threshold is more than one, then, a kind of possible implementation is to choose
The highest image-region of similarity is identified as identification object corresponding image-region in the first image so as to avoid one
Object corresponds to multiple images region in an image.
Alternatively possible implementation is that the position of the corresponding image-region of object is identified in the second image of selected distance
Nearest image-region, as identification object corresponding image-region in the first image.
Since the time interval between the image of monitoring device acquisition is shorter, thus it is same to be identified right in adjacent two images
As mobile distance is shorter, when predicting image-region of the image-region of the identification object in the second image in the first image,
It can be to identify the nearest image-region of the image-region of object in the second image of selected distance, as identification object in the first image
In corresponding image-region, correspond to multiple images region in an image so as to avoid an identification object.
Optionally, each identification object is predicted in the first image after corresponding image-region, for any identification object,
The key point of detection identification object in the image-region of identification object.Identification object is adjusted in the first image according to key point
Corresponding image-region.
For example, set object to be identified as motor vehicle, predict motor vehicle in the first image after corresponding vehicle frame,
The key point in vehicle frame is detected, if including a front window in the key point of monitoring, illustrates the motor vehicle in the vehicle frame
For vehicle frontal, and front vehicle wheel is not included in vehicle frame, then can expand vehicle frame downwards, is made in vehicle frame comprising machine
The front vehicle wheel of motor-car.
Since in Forecasting recognition object, after the image-region in the first other image, detection identifies the key point of object, so
The image-region of identification object is adjusted based on key point afterwards, to make to identify that the image-region of object is more acurrate.
It, can also be to this M in the embodiment of the present invention after finding M image information of the first object to be identified
Image information (image information that the image information and prediction obtained including detection obtains), may further determine that M image letter
It ceases no for identification image (such as pedestrian image or motor vehicle image or non-motor vehicle image).
Due to first judging whether detection image information and forecast image information are images to be recognized, then according to same wait know
The detection image information for having been determined as images to be recognized and forecast image information of other object, determine the identification figure of object to be identified
Picture further increases the subsequent identification image using object to be identified to improve the precision of the identification image of object to be identified
The accuracy filed.
Further, it is contemplated that when the first image is prediction frame image, each image information for identifying object is that prediction obtains
, in order to improve each identification object image information accuracy, the embodiment of the present invention provides to be identified in a kind of prediction frame image
Flow diagram corresponding to the determination method of object, as shown in figure 3, specifically comprising the following steps:
Step 301, determine the first image be detection frame image when, to the first image carry out object detection, determine described in
Corresponding first image information of each first object in first image.
Step 302, according to the corresponding third image information of each second object in third image, predict that each second object exists
Corresponding second image information in first image.
Third image is the adjacent image of the first image of monitoring device acquisition and third image is prediction frame image.
Step 303, meet in the second image information of the second object and the first image information of the first object and impose a condition
When, it determines the second object and the first object is same identification object, and determined according to the first image information and the second image information
The detection image information of same identification object.
In a kind of possible embodiment, in the second image-region of the second object and the first image district of the first object
There are when intersection, determine the second object and the first object for same target in domain.
Illustratively, setting carries out object detection to the first image, determines the first frames images of identification object A, according to the
The corresponding third frames images of each second object in two images predict each second object corresponding second image in the first image
Frame.Set in each second frames images there are second frames images and identification object A the first frames images there are intersections, then should
Identification object in second frames images is determined as identifying object A.
In a kind of possible embodiment, in the locational space and the first object of the second image-region of the second object
When the distance between locational space of first image-region is less than given threshold, determines the second object and the first object is with a pair of
As.
According to the corresponding third image information of each second object in the second image, predict each second object in the first image
In corresponding second image information, the second object the second image information and the first object the first image information meet set
When condition, determines that the second object and the first object are same target, on the one hand avoid same target in an images to be recognized
Two image-regions of middle appearance, while ensure that continuity of the same target in each images to be recognized.
Optionally, after executing above-mentioned steps 203, position letter of the identification object in different frame image can also be recorded
Breath, to be used to form trace information of the identification object under monitoring device.In this way, can be improved it is subsequent to identification object chase after
The convenience of track.
Step 204, the detection image information and forecast image information according to same identification object in different frame image, really
Surely the identification image of object is identified.
There are many modes for determining the identification image of identification object, and a kind of possible implementation is, according to same identification
Detection image information and forecast image information of the object in different frame image choose K image information as identification object
Identify image.Specifically, there are many selection modes of K image information, it can be and chosen according to the quality of each image information
, for example choose the image information of quality preferably (image clearly, image are complete)；Or it is also possible to according to each image information
Shooting angle choose, such as choose shooting angle preferably (positive shooting angle) image information, specifically without limitation.
For example, setting same object to be identified includes 10 image informations in 10 frame images to be recognized, then will be from
8 identification images that 8 quality are preferably used as face object are chosen in this 10 image informations.
Alternatively possible implementation is, according to detection image information of the same identification object in different frame image and
Forecast image information chooses K image information, and the identification image of identification object is generated according to K image information.Specifically, K
The selection mode of each image information is similar with content as described above, is not described in detail herein.
For example, setting same object to be identified includes 10 image informations in 10 frame images to be recognized, then will be from
8 preferable image informations of quality are chosen in this 10 image informations, and are 1 identification image by this 8 image information fusions.
It is alternatively possible to further extract the detection image information and the corresponding category of forecast image information of same identification object
Property information, so that it is determined that identification object attribute information.It is possible to further according to identification object attribute information, to identification
Object is identified or is filed.
Embodiment in order to preferably explain the present invention describes the embodiment of the present invention below with reference to specific implement scene and provides
A kind of video in object determination method, this method can by video object determining device execute.
Setting includes 10 frame images to be recognized in video flowing, and wherein first frame images to be recognized is detection frame image, third
Frame images to be recognized is detection frame image.First frame images to be recognized is detected first, determines first frame images to be recognized
In each identification object the first detection block.For the first detection block for identifying object A in first frame images to be recognized, first to the
It identifies that the key point of object A is detected in one detection block, the first detection block is adjusted according to the key point of detection.Then prediction is known
The first detection block of other object A is in corresponding second prediction block of the second frame images to be recognized.To identification object A in the second prediction block
Key point detected, according to the key point of detection adjust the second face prediction block.The second of Forecasting recognition object A again later
Prediction block corresponding third prediction block in third frame images to be recognized, to the crucial click-through for identifying object A in third prediction block
Row detection adjusts third prediction block according to the key point of detection.Third frame images to be recognized is detected simultaneously, determines third
The third detection block of frame images to be recognized.There are a third detection blocks in the third detection block of setting third frame images to be recognized
There are intersections with the third prediction block of identification object A, then using the third face prediction of third detection block amendment identification object A
Frame.Whether the third prediction block for judging revised identification object A again later is facial image, if so, after forecast value revision
Identify the third prediction block of object A in corresponding 4th prediction block of the 4th frame images to be recognized.And so on, until that cannot predict
Identify object A prediction block in next frame images to be recognized.It sets in 10 frame images to be recognized, corresponding 8 frames of identification object A,
Then using this 8 frames as identification object A identification image, for it is subsequent to identification object A carry out identification or to identification object A into
Row filing.
In addition in the third face detection block of third frame images to be recognized, as the third party with third frame images to be recognized
When face prediction block does not have intersection, such third face detection block is determined as the corresponding Face datection frame of a new face.It is right
Key point in third face detection block is detected, and adjusts third face detection block according to the key point of detection.Then judge
Whether third face detection block is facial image, if so, prediction third face detection block is corresponding in the 4th frame images to be recognized
The 4th face prediction block, and so on, until that cannot predict the face prediction block in next frame images to be recognized.
Based on the same technical idea, the embodiment of the invention provides a kind of determining devices of object in video, such as Fig. 4 institute
Show, which includes acquiring unit 401 and processing unit 402, in which:
Acquiring unit 401, the video to be processed shot within a preset period of time for obtaining monitoring device, it is described to
Handling video includes N frame image；N is more than or equal to 2；
Processing unit 402, if the first image is detection frame image, detects described for being directed to the first image
The type and the corresponding detection image information of each identification object of object are respectively identified in one image；If the first image is prediction frame
Image predicts each identification object described first then according to the corresponding image information of each identification object in the second image
Corresponding forecast image information in image；The first image is any of described N frame image, and second image is institute
It states the adjacent image of the first image and has determined that or predict the corresponding image information of identification object；
The processing unit 402, be also used to according to same detection image information of the identification object in different frame image and
Forecast image information determines the identification image of the identification object.
In one possible implementation, the processing unit 402 is specifically used for:
Determine the first image be detection frame image when, to the first image carry out object detection, determine described in
Corresponding first image information of each first object in first image；
According to the corresponding third image information of each second object in third image, predict each second object described
Corresponding second image information in first image；The third image is the phase of the first image of monitoring device acquisition
Adjacent image and the third image are prediction frame image；
When the second image information of the second object and the first image information of the first object meet and impose a condition, institute is determined
It states the second object and first object is same identification object, and believed according to the first image information and second image
Breath determines the detection image information of the same identification object.
In one possible implementation, the processing unit 402 is specifically used for:
If the first image is detection frame image, the first image is inputted into sorter model, determines described the
The type and the corresponding detection image information of each identification object of object are respectively identified in one image；Class of the classifier for differentiation
Type includes motor vehicle, non-motor vehicle and pedestrian.
In one possible implementation, the processing unit 402 is specifically used for:
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
Identification image of a image information as the identification object；Alternatively,
According to detection image information and forecast image information of the same identification object in different frame image, K is chosen
A image information, and the identification image for identifying object is generated according to the K image information.
The embodiment of the present application also provides a kind of device, which, which has, realizes that object is really in video as described above
Determine the function of method.The function can execute corresponding software realization by hardware, in a kind of possible design, the device packet
It includes: processor, transceiver, memory；The memory is for storing computer executed instructions, and the transceiver is for realizing the device
It is communicated with other communication entities, which is connect with the memory by the bus, when the apparatus is operative, the processing
Device executes the computer executed instructions of memory storage, so that the device executes in video as described above object really
Determine method.
The embodiment of the present invention also provides a kind of computer storage medium, stores software program in the storage medium, this is soft
Part program realizes view described in above-mentioned various possible implementations when being read and executed by one or more processors
The determination method of object in frequency.
The embodiment of the present invention also provides a kind of computer program product comprising instruction, when run on a computer,
So that computer executes the determination method of object in video described in above-mentioned various possible implementations.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (10)
1. a kind of determination method of object in video, which is characterized in that the described method includes:
The video to be processed that monitoring device is shot within a preset period of time is obtained, the video to be processed includes N frame image；
N is more than or equal to 2；
It detects if the first image is detection frame image for the first image and respectively identifies object in the first image
Type and the corresponding detection image information of each identification object；If the first image is prediction frame image, according to the second image
In the corresponding image information of each identification object, predict each identification object corresponding forecast image in the first image
Information；The first image is any of described N frame image, and second image is the adjacent image of the first image
And have determined that or predict the corresponding image information of identification object；
According to detection image information and forecast image information of the same identification object in different frame image, the identification pair is determined
The identification image of elephant.
2. the method according to claim 1, wherein if the first image is detection frame image, it is determined that each
Identify the corresponding detection image information of object, comprising:
When determining the first image is detection frame image, object detection is carried out to the first image, determines described first
Corresponding first image information of each first object in image；
According to the corresponding third image information of each second object in third image, predict each second object described first
Corresponding second image information in image；The third image is the neighbor map of the first image of monitoring device acquisition
Picture and the third image are prediction frame image；
When the second image information of the second object and the first image information of the first object meet and impose a condition, described the is determined
Two objects and first object are same identification object, and true according to the first image information and second image information
The detection image information of the fixed same identification object.
3. the method according to claim 1, wherein if the first image is detection frame image, it is determined that institute
State the type that object is respectively identified in the first image and the corresponding detection image information of each identification object, comprising:
If the first image is detection frame image, the first image is inputted into sorter model, determines first figure
The type and the corresponding detection image information of each identification object of object are respectively identified as in；Type packet of the classifier for differentiation
Include motor vehicle, non-motor vehicle and pedestrian.
4. according to the method in any one of claims 1 to 3, which is characterized in that according to same identification object in different frame
Detection image information and forecast image information in image determine the identification image of the identification object, comprising:
According to detection image information and forecast image information of the same identification object in different frame image, K figure is chosen
Identification image as information as the identification object；Alternatively,
According to detection image information and forecast image information of the same identification object in different frame image, K figure is chosen
The identification image for identifying object is generated as information, and according to the K image information.
5. the determining device of object in a kind of video, which is characterized in that described device includes:
Acquiring unit, the video to be processed shot within a preset period of time for obtaining monitoring device, the view to be processed
Frequency includes N frame image；N is more than or equal to 2；
Processing unit, if the first image is detection frame image, detects in the first image for being directed to the first image
The type of each identification object and the corresponding detection image information of each identification object；If the first image is prediction frame image,
According to the corresponding image information of each identification object in the second image, predict that each identification object is right in the first image
The forecast image information answered；The first image is any of described N frame image, and second image is first figure
As adjacent image and have determined that or predict the corresponding image information of identification object；
The processing unit is also used to detection image information and forecast image according to same identification object in different frame image
Information determines the identification image of the identification object.
6. device according to claim 5, which is characterized in that the processing unit is specifically used for:
When determining the first image is detection frame image, object detection is carried out to the first image, determines described first
Corresponding first image information of each first object in image；
According to the corresponding third image information of each second object in third image, predict each second object described first
Corresponding second image information in image；The third image is the neighbor map of the first image of monitoring device acquisition
Picture and the third image are prediction frame image；
When the second image information of the second object and the first image information of the first object meet and impose a condition, described the is determined
Two objects and first object are same identification object, and true according to the first image information and second image information
The detection image information of the fixed same identification object.
7. device according to claim 5, which is characterized in that the processing unit is specifically used for:
If the first image is detection frame image, the first image is inputted into sorter model, determines first figure
The type and the corresponding detection image information of each identification object of object are respectively identified as in；Type packet of the classifier for differentiation
Include motor vehicle, non-motor vehicle and pedestrian.
8. device according to any one of claims 5 to 7, which is characterized in that the processing unit is specifically used for:
According to detection image information and forecast image information of the same identification object in different frame image, K figure is chosen
Identification image as information as the identification object；Alternatively,
According to detection image information and forecast image information of the same identification object in different frame image, K figure is chosen
The identification image for identifying object is generated as information, and according to the K image information.
9. a kind of computer readable storage medium, which is characterized in that the storage medium is stored with instruction, when described instruction is being counted
When being run on calculation machine, so that computer realizes method described in any one of perform claim requirement 1 to 4.
10. a kind of computer equipment characterized by comprising
Memory, for storing program instruction；
Processor, for calling the program instruction stored in the memory, according to acquisition program execute as claim 1 to
Method described in any claim in 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811648094.8A CN109800685A (en) | 2018-12-29 | 2018-12-29 | The determination method and device of object in a kind of video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811648094.8A CN109800685A (en) | 2018-12-29 | 2018-12-29 | The determination method and device of object in a kind of video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109800685A true CN109800685A (en) | 2019-05-24 |
Family
ID=66558125
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811648094.8A Pending CN109800685A (en) | 2018-12-29 | 2018-12-29 | The determination method and device of object in a kind of video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109800685A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110675433A (en) * | 2019-10-31 | 2020-01-10 | 北京达佳互联信息技术有限公司 | Video processing method and device, electronic equipment and storage medium |
| CN114466218A (en) * | 2022-02-18 | 2022-05-10 | 广州方硅信息技术有限公司 | Live video character tracking method, device, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631419A (en) * | 2015-12-24 | 2016-06-01 | 浙江宇视科技有限公司 | Face recognition method and device |
| CN106845357A (en) * | 2016-12-26 | 2017-06-13 | 银江股份有限公司 | A kind of video human face detection and recognition methods based on multichannel network |
| CN108876812A (en) * | 2017-11-01 | 2018-11-23 | 北京旷视科技有限公司 | Image processing method, device and equipment for object detection in video |
| CN108986138A (en) * | 2018-05-24 | 2018-12-11 | 北京飞搜科技有限公司 | Method for tracking target and equipment |
| CN109035304A (en) * | 2018-08-07 | 2018-12-18 | 北京清瑞维航技术发展有限公司 | Method for tracking target, calculates equipment and device at medium |
-
2018
- 2018-12-29 CN CN201811648094.8A patent/CN109800685A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631419A (en) * | 2015-12-24 | 2016-06-01 | 浙江宇视科技有限公司 | Face recognition method and device |
| CN106845357A (en) * | 2016-12-26 | 2017-06-13 | 银江股份有限公司 | A kind of video human face detection and recognition methods based on multichannel network |
| CN108876812A (en) * | 2017-11-01 | 2018-11-23 | 北京旷视科技有限公司 | Image processing method, device and equipment for object detection in video |
| CN108986138A (en) * | 2018-05-24 | 2018-12-11 | 北京飞搜科技有限公司 | Method for tracking target and equipment |
| CN109035304A (en) * | 2018-08-07 | 2018-12-18 | 北京清瑞维航技术发展有限公司 | Method for tracking target, calculates equipment and device at medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110675433A (en) * | 2019-10-31 | 2020-01-10 | 北京达佳互联信息技术有限公司 | Video processing method and device, electronic equipment and storage medium |
| US11450027B2 (en) | 2019-10-31 | 2022-09-20 | Beijing Dajia Internet Information Technologys Co., Ltd. | Method and electronic device for processing videos |
| CN114466218A (en) * | 2022-02-18 | 2022-05-10 | 广州方硅信息技术有限公司 | Live video character tracking method, device, equipment and storage medium |
| CN114466218B (en) * | 2022-02-18 | 2024-04-23 | 广州方硅信息技术有限公司 | Live video character tracking method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11854136B2 (en) | Monitoring a scene to analyze an event using a plurality of image streams | |
| US20200364467A1 (en) | Method and device for detecting illegal parking, and electronic device | |
| CN110753892A (en) | Method and system for instant object tagging via cross-modality verification in autonomous vehicles | |
| CN110869559A (en) | Method and system for integrated global and distributed learning in autonomous vehicles | |
| CN110753953A (en) | Method and system for object-centric stereo vision in autonomous vehicles via cross-modality verification | |
| CN103208008A (en) | Fast adaptation method for traffic video monitoring target detection based on machine vision | |
| CN111160243A (en) | Passenger flow volume statistical method and related product | |
| CN109325429A (en) | A kind of method, apparatus, storage medium and the terminal of linked character data | |
| CN109815842A (en) | A kind of method and device of the attribute information of determining object to be identified | |
| CN109784254A (en) | A kind of method, apparatus and electronic equipment of rule-breaking vehicle event detection | |
| CN109800675A (en) | A kind of method and device of the identification image of determining face object | |
| CN105608209A (en) | Video labeling method and video labeling device | |
| CN109872541A (en) | A kind of information of vehicles analysis method and device | |
| CN109800678A (en) | The attribute determining method and device of object in a kind of video | |
| CN109800685A (en) | The determination method and device of object in a kind of video | |
| CN109800684A (en) | The determination method and device of object in a kind of video | |
| CN112738725B (en) | Real-time identification method, device, equipment and medium for target crowd in semi-closed area | |
| KR20170095599A (en) | System and method for video searching | |
| CN111064924B (en) | Video monitoring method and system based on artificial intelligence | |
| CN113160279A (en) | Method and device for detecting abnormal behaviors of pedestrians in subway environment | |
| CN109784238A (en) | A kind of method and device of determining object to be identified | |
| CN114078319A (en) | Method and device for detecting potential hazard site of traffic accident | |
| CN109508703A (en) | A kind of face in video determines method and device | |
| CN109740525A (en) | The determination method and device of object in a kind of video | |
| CN109740518A (en) | The determination method and device of object in a kind of video |
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 |
Application publication date: 20190524 |
|
| RJ01 | Rejection of invention patent application after publication |