CN109815872A - Cheating method for detecting area, device, equipment and storage medium - Google Patents
Cheating method for detecting area, device, equipment and storage medium Download PDFInfo
- Publication number
- CN109815872A CN109815872A CN201910042318.9A CN201910042318A CN109815872A CN 109815872 A CN109815872 A CN 109815872A CN 201910042318 A CN201910042318 A CN 201910042318A CN 109815872 A CN109815872 A CN 109815872A
- Authority
- CN
- China
- Prior art keywords
- frame
- cheating
- hot spot
- region
- spot region
- 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 cheating method for detecting area, device, equipment and storage mediums, and the pixel coordinate position of person head in each frame picture by obtaining video to be identified generates number of people coordinate data according to the pixel coordinate position;The hot spot region frame of each frame picture is obtained according to the number of people coordinate data;Each hot spot region frame is iterated cluster, obtain the overlapping density of each hot spot region frame, the corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region, preventing manually invigilator, there are internal staff's malpractices situations, save human cost, improve the accuracy rate of cheating region detection, improve the detection speed and efficiency in cheating region, it ensure that the fair and just of examination, it does not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility, the user experience is improved.
Description
Technical field
The present invention relates to field of video monitoring more particularly to a kind of cheating method for detecting area, device, equipment and storage to be situated between
Matter.
Background technique
Detection cheating or the mode of behavior are usually to pass through manually to invigilate, or move a certain of target person at present
Static single frames acquisition is done as behavior, is differentiated after collected frame picture progress statistical learning is gone out model, but because
It is more to differentiate identification maneuver, model is caused to tend to complicate further, calculation amount is huger, needs to rely on very big data volume fortune
Calculation ability, often arithmetic speed is partially slow, and only by determining cheating region when former frame picture, Detection accuracy is not high, and
And the existing action behavior by extracting target person carries out this mode of cheating analysis, many toward contact needs dependence
The image information and sensor information that hardware sensor obtains, have that noise is big, and in more complicated monitoring ring
Real-time is not also high in border, has delay.
Summary of the invention
The main purpose of the present invention is to provide a kind of cheating method for detecting area, device, equipment and storage mediums, it is intended to
Solve in the prior art by extract the action behavior of target person carry out Detection accuracy and efficiency existing for cheating analysis compared with
It is low, and the technical problem of real-time difference.
To achieve the above object, the present invention provides a kind of cheating method for detecting area, the cheating method for detecting area packet
Include following steps:
The pixel coordinate position for obtaining person head in each frame picture of video to be identified, according to the pixel coordinate position
Generate number of people coordinate data;
The hot spot region frame of each frame picture is obtained according to the number of people coordinate data;
Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame;
The corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region.
Preferably, described that each hot spot region frame is iterated cluster, the overlapping density of each hot spot region frame is obtained, specifically
Include:
Each frame picture is overlapped, the accumulative overlapping number of each hot spot region frame is counted;
The overlapping density of each hot spot region frame is determined according to the accumulative overlapping number.
Preferably, described that each frame picture is overlapped, the accumulative overlapping number of each hot spot region frame is counted, it is specific to wrap
It includes:
Continuous each frame picture is successively overlapped, so that each hot spot region frame in each frame picture is merged;
The regional ensemble for obtaining fused each hot spot region frame, obtains each hot spot region frame from the regional ensemble
Accumulative overlapping number, the regional ensemble are obtained by following formula:
Wherein, n is iteration threshold, i.e. overlapping number between frame and frame, and i is current frame number, HiFor the hot spot of present frame
Regional frame, W are regional ensemble.
Preferably, in each frame picture for obtaining video to be identified person head pixel coordinate position, according to described
Pixel coordinate position generates number of people coordinate data, specifically includes:
The video to be identified is identified, determines the person head mesh in each frame picture of the video to be identified
Mark;
The corresponding pixel coordinate position of person head target and person head quantity in each frame picture are obtained, each pixel is sat
Cursor position and person head quantity are as number of people coordinate data.
Preferably, the hot spot region frame that each frame picture is obtained according to the number of people coordinate data, specifically includes:
The corresponding target person headers box in each target person head is determined according to the number of people coordinate data;
Obtain the spacing in each frame picture between each target person headers box;
The spacing is less than the target person headers box of pre-determined distance threshold value as hot spot region frame.
Preferably, described that the corresponding target person headers box in each target person head is determined according to the number of people coordinate data
Before, the cheating method for detecting area further include:
From the starting number of people coordinate data and last frame picture extracted in the number of people coordinate data in first frame picture
In final number of people coordinate data;
The starting number of people coordinate data is compared with the final number of people coordinate data, obtains person head displacement
Parameter, the person head displacement parameter are that the history of the pixel coordinate position on everyone head in first frame picture is displaced number
According to;
Determine that person head displacement is less than the person head of preset displacement threshold value according to the person head displacement parameter
Target person head.
Preferably, the corresponding region of the maximum hot spot region frame of density that will be overlapped is used as after cheating region, described
Cheating method for detecting area further include:
The corresponding destination image data in the cheating region is obtained, the destination image data is substituting to default erroneous detection mould
In type, erroneous detection result is generated;
It is that nothing is mistaken in the erroneous detection result, it is raw according to the prewarning area using the cheating region as prewarning area
Monitoring center is sent at warning information, and by the warning information.
In addition, to achieve the above object, the present invention also proposes a kind of cheating equipment for area detection equipment, the cheating region detection
Equipment includes: the cheating region inspection that memory, processor and being stored in can be run on the memory and on the processor
The step of ranging sequence, the cheating region detector is arranged for carrying out cheating method for detecting area as described above.
In addition, to achieve the above object, the present invention also proposes a kind of storage medium, cheating is stored on the storage medium
Region detector, the cheating region detector realize cheating region detection side as described above when being executed by processor
The step of method.
In addition, to achieve the above object, the present invention also provides a kind of cheating regional detection device, the cheating region detection
Device includes: data acquisition module, hot spot region determining module, density acquisition module and cheating area determination module;
Wherein, the data acquisition module, the pixel of person head is sat in each frame picture for obtaining video to be identified
Cursor position generates number of people coordinate data according to the pixel coordinate position;
The hot spot region determining module, for obtaining the hot spot region of each frame picture according to the number of people coordinate data
Frame;
The density obtains module and obtains the weight of each hot spot region frame for each hot spot region frame to be iterated cluster
Folded density;
The cheating area determination module, for the corresponding region of the maximum hot spot region frame of density will to be overlapped as cheating
Region.
Cheating method for detecting area proposed by the present invention passes through person head in each frame picture of acquisition video to be identified
Pixel coordinate position generates number of people coordinate data according to the pixel coordinate position;It is obtained according to the number of people coordinate data each
The hot spot region frame of frame picture;Each hot spot region frame is iterated cluster, the overlapping density of each hot spot region frame is obtained, will weigh
The corresponding region of the maximum hot spot region frame of density is folded as cheating region, preventing manually invigilator, there are internal staff's malpractices feelings
Condition saves human cost, improves the accuracy rate of cheating region detection, improves the detection speed and efficiency in cheating region,
It ensure that the fair and just of examination, do not need that the test object of each frame image is marked and is associated with, versatility and flexibility
By force, the user experience is improved.
Detailed description of the invention
Fig. 1 is the device structure schematic diagram for the hardware running environment that the embodiment of the present invention is related to;
Fig. 2 is the flow diagram of present invention cheating method for detecting area first embodiment;
Fig. 3 is the flow diagram of present invention cheating method for detecting area second embodiment;
Fig. 4 is the flow diagram of present invention cheating method for detecting area 3rd embodiment;
Fig. 5 is the functional block diagram of present invention cheating regional detection device first embodiment.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
The solution of the embodiment of the present invention is mainly: personnel in each frame picture of the present invention by obtaining video to be identified
The pixel coordinate position on head generates number of people coordinate data according to the pixel coordinate position;According to the number of people coordinate data
Obtain the hot spot region frame of each frame picture;Each hot spot region frame is iterated cluster, the overlapping for obtaining each hot spot region frame is close
Degree will be overlapped the corresponding region of the maximum hot spot region frame of density as cheating region, and preventing manually invigilator, there are internal staff
Malpractices situation, saves human cost, improve cheating region detection accuracy rate, improve cheating region detection speed and
Efficiency ensure that the fair and just of examination, not need that the test object of each frame image is marked and is associated with, versatility and spirit
Active strong, the user experience is improved, solves and carries out cheating analysis by extracting the action behavior of target person in the prior art
Existing Detection accuracy and efficiency are lower, and the technical problem of real-time difference.
Referring to Fig.1, Fig. 1 is the device structure schematic diagram for the hardware running environment that the embodiment of the present invention is related to.
As shown in Figure 1, the equipment may include: display 1000, processor 1001, such as central processing unit (Central
Processing Unit, CPU), communication bus 1002, user interface 1003, network interface 1004, memory 1005.Wherein,
Communication bus 1002 is for realizing the connection communication between these components.User interface 1003 may include that the wired of standard connects
Mouth, wireless interface.Network interface 1004 optionally may include standard wireline interface and wireless interface (such as Wireless Fidelity
(WIreless-FIdelity, WI-FI) interface).Memory 1005 can be the random access memory (Random of high speed
Access Memory, RAM) memory, it is also possible to stable memory (Non-volatile Memory, NVM), such as magnetic
Disk storage.Memory 1005 optionally can also be the storage device independently of aforementioned processor 1001.
It, can be with it will be understood by those skilled in the art that device structure shown in Fig. 1 does not constitute the restriction to the equipment
Including perhaps combining certain components or different component layouts than illustrating more or fewer components.
As shown in Figure 1, as may include operating system, network communication mould in a kind of memory 1005 of storage medium
Block, user terminal interface module and cheating region detector.
Present invention cheating equipment for area detection equipment calls the cheating region stored in memory 1005 inspection by processor 1001
Ranging sequence, and execute following operation:
The pixel coordinate position for obtaining person head in each frame picture of video to be identified, according to the pixel coordinate position
Generate number of people coordinate data;
The hot spot region frame of each frame picture is obtained according to the number of people coordinate data;
Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame;
The corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
Each frame picture is overlapped, the accumulative overlapping number of each hot spot region frame is counted;
The overlapping density of each hot spot region frame is determined according to the accumulative overlapping number.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
Continuous each frame picture is successively overlapped, so that each hot spot region frame in each frame picture is merged;
The regional ensemble for obtaining fused each hot spot region frame, obtains each hot spot region frame from the regional ensemble
Accumulative overlapping number, the regional ensemble are obtained by following formula:
Wherein, n is iteration threshold, i.e. overlapping number between frame and frame, and i is current frame number, HiFor the hot spot of present frame
Regional frame, W are regional ensemble.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
The video to be identified is identified, determines the person head mesh in each frame picture of the video to be identified
Mark;
The corresponding pixel coordinate position of person head target and person head quantity in each frame picture are obtained, each pixel is sat
Cursor position and person head quantity are as number of people coordinate data.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
The corresponding target person headers box in each target person head is determined according to the number of people coordinate data;
Obtain the spacing in each frame picture between each target person headers box;
The spacing is less than the target person headers box of pre-determined distance threshold value as hot spot region frame.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
From the starting number of people coordinate data and last frame picture extracted in the number of people coordinate data in first frame picture
In final number of people coordinate data;
The starting number of people coordinate data is compared with the final number of people coordinate data, obtains person head displacement
Parameter, the person head displacement parameter are that the history of the pixel coordinate position on everyone head in first frame picture is displaced number
According to;
Determine that person head displacement is less than the person head of preset displacement threshold value according to the person head displacement parameter
Target person head.
Further, processor 1001 can call the cheating region detector stored in memory 1005, also execute
It operates below:
The corresponding destination image data in the cheating region is obtained, the destination image data is substituting to default erroneous detection mould
In type, erroneous detection result is generated;
It is that nothing is mistaken in the erroneous detection result, it is raw according to the prewarning area using the cheating region as prewarning area
Monitoring center is sent at warning information, and by the warning information.
The present embodiment through the above scheme, passes through the pixel coordinate of person head in each frame picture of acquisition video to be identified
Position generates number of people coordinate data according to the pixel coordinate position;Each frame picture is obtained according to the number of people coordinate data
Hot spot region frame;Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame, most by overlapping density
The corresponding region of big hot spot region frame prevents from manually invigilating there are internal staff's malpractices situation, save as cheating region
Human cost improves the accuracy rate of cheating region detection, improves the detection speed and efficiency in cheating region, ensure that examination
It is fair and just, do not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility improve use
Family experience.
Based on above-mentioned hardware configuration, present invention cheating method for detecting area embodiment is proposed.
It is the flow diagram of present invention cheating method for detecting area first embodiment referring to Fig. 2, Fig. 2.
In the first embodiment, the cheating method for detecting area the following steps are included:
Step S10, the pixel coordinate position for obtaining person head in each frame picture of video to be identified, according to the pixel
Coordinate position generates number of people coordinate data.
It should be noted that the video to be identified is the pre-set video for carrying out cheating region detection, institute
Stating video to be identified can be the video prerecorded, and be also possible to the video of real-time monitoring acquisition, in the present embodiment institute
Stating video to be identified is the video that real-time monitoring obtains, and can also be that this is not added in the video of other forms, the present embodiment certainly
With limitation;The personnel that the pixel coordinate position of the person head occurs in the frame picture for each frame of the video to be identified
Head, the i.e. specific location of pixels in the frame picture of each frame of person head, and then corresponding number of people coordinate data is generated,
The number of people coordinate data reflects that person head is in the specific location of corresponding frame picture in each frame picture.
Further, the step S10 specifically includes the following steps:
The video to be identified is identified, determines the person head mesh in each frame picture of the video to be identified
Mark;
The corresponding pixel coordinate position of person head target and person head quantity in each frame picture are obtained, each pixel is sat
Cursor position and person head quantity are as number of people coordinate data.
It should be understood that the video to be identified is identified, it can be by grabbing the personnel in each frame picture
Head pixel determines the person head target in each frame picture of the video to be identified, and then records everyone head mesh
Mark corresponding pixel coordinate position and count everyone head target by the corresponding personnel of person head target in each frame picture
Head quantity, thus using each pixel coordinate position and person head quantity as number of people coordinate data, it is ensured that subsequent hot spot
The quick determination of regional frame improves the detection speed and efficiency in cheating region.
Step S20, the hot spot region frame of each frame picture is obtained according to the number of people coordinate data.
It is understood that the hot spot region frame is the corresponding region of number of people coordinate data with larger cheating suspicion
Frame can obtain the hot spot region of each frame picture, i.e. locking has compared with your writing by analyzing the number of people coordinate data
The hot spot region of disadvantage suspicion.
Step S30, each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame.
People is calculated it should be understood that each hot spot region frame is iterated cluster and be overlapped to each hot spot region frame
The density that member head covers mutually, and then each hot spot region frame can be screened, the overlapping for obtaining each hot spot region frame is close
Degree.
Step S40, the corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region.
It is understood that the overlapping maximum hot spot region frame of density is the maximum hot spot region frame of cheating suspicion, it will
The corresponding region of the maximum hot spot region frame of density is overlapped as cheating region, the accuracy rate of cheating region detection is improved, mentions
The high detection speed and efficiency in cheating region.
Further, after the step S40, the cheating method for detecting area is further comprising the steps of:
The corresponding destination image data in the cheating region is obtained, the destination image data is substituting to default erroneous detection mould
In type, erroneous detection result is generated;
It is that nothing is mistaken in the erroneous detection result, it is raw according to the prewarning area using the cheating region as prewarning area
Monitoring center is sent at warning information, and by the warning information.
It is understood that after cheating region has been determined, it can be by the default erroneous detection model to the cheating area
The case where domain is filtered erroneous detection, excludes erroneous detection, the default erroneous detection model are pre-set for carrying out to target image
The model that erroneous detection excludes, the default erroneous detection model can be obtained by lot of experimental data training, be also possible to pass through technology
The empirically determined model of the regular job of personnel, can also be determining by other means certainly is model, and the present embodiment is to this
It is without restriction;The erroneous detection result in the cheating region is being determined as without mistaking, using the cheating region as prewarning area,
Warning information is generated according to the prewarning area, and the warning information is sent to monitoring center, thus notice monitoring in time
The staff at center carries out the respective handling of cheating, and the fair and just of examination has been effectively ensured.
The present embodiment through the above scheme, passes through the pixel coordinate of person head in each frame picture of acquisition video to be identified
Position generates number of people coordinate data according to the pixel coordinate position;Each frame picture is obtained according to the number of people coordinate data
Hot spot region frame;Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame, most by overlapping density
The corresponding region of big hot spot region frame prevents from manually invigilating there are internal staff's malpractices situation, save as cheating region
Human cost improves the accuracy rate of cheating region detection, improves the detection speed and efficiency in cheating region, ensure that examination
It is fair and just, do not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility improve use
Family experience.
Further, Fig. 3 is the flow diagram of present invention cheating method for detecting area second embodiment, as shown in figure 3,
Present invention cheating method for detecting area second embodiment is proposed based on first embodiment, in the present embodiment, the step S30,
Specifically includes the following steps:
Step S31, each frame picture is overlapped, counts the accumulative overlapping number of each hot spot region frame.
It should be noted that being overlapped to each frame picture, i.e., next frame picture is subjected to Chong Die melt with previous frame picture
It closes, and then obtains the hot spot region frame number Chong Die with the hot spot region frame in previous frame picture in next frame picture, successively obtain
The overlapping cases for obtaining each hot spot region frame in each frame picture, that is, count the accumulative overlapping number of each hot spot region frame.
Further, the step S31, specifically includes the following steps:
Continuous each frame picture is successively overlapped, so that each hot spot region frame in each frame picture is merged;
The regional ensemble for obtaining fused each hot spot region frame, obtains each hot spot region frame from the regional ensemble
Accumulative overlapping number, the regional ensemble are obtained by following formula:
Wherein, n is iteration threshold, i.e. overlapping number between frame and frame, and i is current frame number, HiFor the hot spot of present frame
Regional frame, W are regional ensemble.
It should be understood that the video to be identified can be obtained by being successively overlapped to continuous each frame picture
Each frame picture hot spot region overlapping cases, i.e., each hot spot region frame in each frame picture carries out fused fusion feelings
Condition, i.e., obtain the accumulative overlapping number of each hot spot region frame in each regional ensemble, and the accumulative overlapping number reflects each frame picture
The corresponding personnel in hot spot region cheating suspicion size, the corresponding personnel in the accumulative more big then hot spot region of overlapping number
Cheating suspicion it is bigger, W is regional ensemble, by the accumulative overlapping time for obtaining each hot spot region frame from the regional ensemble
Number can learn the most hot spot region of overlapping number for cheating region.
Step S32, the overlapping density of each hot spot region frame is determined according to the accumulative overlapping number.
It is understood that the size of the accumulative overlapping number reflects the size of the overlapping density of each hot spot region frame,
The overlapping maximum region of density then shows that being overlapped the corresponding personnel in the maximum region of density, there is the short distance of duration to connect
The case where touching, then corresponding personnel's cheating suspicion is larger in the overlapping maximum region of density at this time, then will be overlapped density most at this time
The corresponding region of big hot spot region frame is as cheating region.
The present embodiment through the above scheme, by being overlapped to each frame picture, counts the accumulative weight of each hot spot region frame
Folded number;The overlapping density that each hot spot region frame is determined according to the accumulative overlapping number, will be overlapped the maximum hot zone of density
Frame corresponding region in domain prevents from manually invigilating there are internal staff's malpractices situation, saves human cost, mention as cheating region
The accuracy rate for having risen cheating region detection improves the detection speed and efficiency in cheating region, ensure that the fair and just of examination,
It does not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility, the user experience is improved.
Further, Fig. 4 is the flow diagram of present invention cheating method for detecting area 3rd embodiment, as shown in figure 4,
Present invention cheating method for detecting area 3rd embodiment is proposed based on second embodiment, in the present embodiment, the step S20 tool
Body the following steps are included:
Step S21, the corresponding target person headers box in each target person head is determined according to the number of people coordinate data.
It should be noted that determining the corresponding target person head in each target person head according to the number of people coordinate data
Frame, i.e., the person head frame of the region setting rule formed the corresponding pixel in everyone head by number of people coordinate data as
Target person headers box, the target person headers box can be square or rectangular, can also be round person head frame,
Certainly it can also be that the person head frame of other forms, the present embodiment are without restriction to this.
Before the step S21, the cheating method for detecting area is further comprising the steps of:
From the starting number of people coordinate data and last frame picture extracted in the number of people coordinate data in first frame picture
In final number of people coordinate data;
The starting number of people coordinate data is compared with the final number of people coordinate data, obtains person head displacement
Parameter, the person head displacement parameter are that the history of the pixel coordinate position on everyone head in first frame picture is displaced number
According to;
Determine that person head displacement is less than the person head of preset displacement threshold value according to the person head displacement parameter
Target person head.
It is understood that first frame picture and last frame picture can be extracted from the number of people coordinate data, into
And the starting number of people coordinate data in first frame picture and the final number of people coordinate data in last frame picture can be carried out
It compares, obtains person head displacement data, the person head displacement parameter is determined that person head displacement is less than preset displacement
The person head of threshold value is target person head, and the corresponding person head displacement of person head displacement parameter is greater than or equal in advance
It, can be with if the person head of displacement threshold value is considered person head of incoherent personnel, such as the person head of pedestrian etc.
Screening further is filtered to number of people coordinate data, the accuracy rate of cheating region detection is further improved, improves cheating
The detection speed and efficiency in region.
Step S22, the spacing in each frame picture between each target person headers box is obtained.
It is understood that the spacing between each target person headers box can reflect between person head away from
From obtaining the spacing in each frame picture between each target person headers box can be central point by each target person headers box
Distance calculate, be also possible to the same relative position in each target person headers box apart from the distance between collection point
It calculates, can also be and otherwise calculated certainly, the present embodiment is without restriction to this.
Step S23, the spacing is less than the target person headers box of pre-determined distance threshold value as hot spot region frame.
It should be understood that the pre-determined distance threshold value be it is pre-set for judge the target person headers box it
Between distance whether excessively close threshold value can be according to comparison result by by the spacing and the pre-determined distance threshold value comparison
It determines whether target person headers box is used as hot spot region frame, and then can effectively be in the target of opposite normal range (NR) to spacing
Person head frame is filtered, using spacing be less than pre-determined distance threshold value target person headers box as hot spot region frame, i.e., between
There is larger cheating suspicion away from the corresponding personnel of target person headers box of pre-determined distance therewith are less than, to improve cheating area
The accuracy rate of domain detection.
The present embodiment through the above scheme, by determining that each target person head is corresponding according to the number of people coordinate data
Target person headers box;Obtain the spacing in each frame picture between each target person headers box;By the spacing be less than it is default away from
Target person headers box from threshold value is prevented from manually invigilating there are internal staff's malpractices situation, be saved as hot spot region frame
Human cost improves the accuracy rate of cheating region detection, improves the detection speed and efficiency in cheating region, ensure that examination
It is fair and just, do not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility improve use
Family experience.
Based on the embodiment of above-mentioned cheating method for detecting area, the present invention further provides a kind of cheating region detection dresses
It sets.
It is the functional block diagram of present invention cheating regional detection device first embodiment referring to Fig. 5, Fig. 5.
The present invention practises fraud in regional detection device first embodiment, which includes: data acquisition mould
Block 10, hot spot region determining module 20, density obtain module 30 and cheating area determination module 40;
Wherein, the data acquisition module 10, the pixel of person head in each frame picture for obtaining video to be identified
Coordinate position generates number of people coordinate data according to the pixel coordinate position;
The hot spot region determining module 20, for obtaining the hot spot region of each frame picture according to the number of people coordinate data
Frame;
The density obtains module 30 and obtains each hot spot region frame for each hot spot region frame to be iterated cluster
It is overlapped density;
The cheating area determination module 40, for the corresponding region of the maximum hot spot region frame of density will to be overlapped as work
Disadvantage region.
Wherein, practise fraud regional detection device each Implement of Function Module the step of can refer to the present invention cheating region detection
Each embodiment of method, details are not described herein again.
In addition, the embodiment of the present invention also proposes a kind of storage medium, cheating region detection is stored on the storage medium
Program realizes following operation when the cheating region detector is executed by processor:
The pixel coordinate position for obtaining person head in each frame picture of video to be identified, according to the pixel coordinate position
Generate number of people coordinate data;
The hot spot region frame of each frame picture is obtained according to the number of people coordinate data;
Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame;
The corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region.
Further, following operation is also realized when the cheating region detector is executed by processor:
Each frame picture is overlapped, the accumulative overlapping number of each hot spot region frame is counted;
The overlapping density of each hot spot region frame is determined according to the accumulative overlapping number.
Further, following operation is also realized when the cheating region detector is executed by processor:
Continuous each frame picture is successively overlapped, so that each hot spot region frame in each frame picture is merged;
The regional ensemble for obtaining fused each hot spot region frame, obtains each hot spot region frame from the regional ensemble
Accumulative overlapping number, the regional ensemble are obtained by following formula:
Wherein, n is iteration threshold, i.e. overlapping number between frame and frame, and i is current frame number, HiFor the hot spot of present frame
Regional frame, W are regional ensemble.
Further, following operation is also realized when the cheating region detector is executed by processor:
The video to be identified is identified, determines the person head mesh in each frame picture of the video to be identified
Mark;
The corresponding pixel coordinate position of person head target and person head quantity in each frame picture are obtained, each pixel is sat
Cursor position and person head quantity are as number of people coordinate data.
Further, following operation is also realized when the cheating region detector is executed by processor:
The corresponding target person headers box in each target person head is determined according to the number of people coordinate data;
Obtain the spacing in each frame picture between each target person headers box;
The spacing is less than the target person headers box of pre-determined distance threshold value as hot spot region frame.
Further, following operation is also realized when the cheating region detector is executed by processor:
From the starting number of people coordinate data and last frame picture extracted in the number of people coordinate data in first frame picture
In final number of people coordinate data;
The starting number of people coordinate data is compared with the final number of people coordinate data, obtains person head displacement
Parameter, the person head displacement parameter are that the history of the pixel coordinate position on everyone head in first frame picture is displaced number
According to;
Determine that person head displacement is less than the person head of preset displacement threshold value according to the person head displacement parameter
Target person head.
Further, following operation is also realized when the cheating region detector is executed by processor:
The corresponding destination image data in the cheating region is obtained, the destination image data is substituting to default erroneous detection mould
In type, erroneous detection result is generated;
It is that nothing is mistaken in the erroneous detection result, it is raw according to the prewarning area using the cheating region as prewarning area
Monitoring center is sent at warning information, and by the warning information.
The present embodiment through the above scheme, passes through the pixel coordinate of person head in each frame picture of acquisition video to be identified
Position generates number of people coordinate data according to the pixel coordinate position;Each frame picture is obtained according to the number of people coordinate data
Hot spot region frame;Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame, most by overlapping density
The corresponding region of big hot spot region frame prevents from manually invigilating there are internal staff's malpractices situation, save as cheating region
Human cost improves the accuracy rate of cheating region detection, improves the detection speed and efficiency in cheating region, ensure that examination
It is fair and just, do not need that the test object of each frame image is marked and is associated with, versatility and strong flexibility improve use
Family experience.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or the device that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or device institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, method of element, article or device.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of cheating method for detecting area, which is characterized in that the cheating method for detecting area includes:
The pixel coordinate position for obtaining person head in each frame picture of video to be identified is generated according to the pixel coordinate position
Number of people coordinate data;
The hot spot region frame of each frame picture is obtained according to the number of people coordinate data;
Each hot spot region frame is iterated cluster, obtains the overlapping density of each hot spot region frame;
The corresponding region of the maximum hot spot region frame of density will be overlapped as cheating region.
2. cheating method for detecting area as described in claim 1, which is characterized in that described to be iterated each hot spot region frame
Cluster, obtains the overlapping density of each hot spot region frame, specifically includes:
Each frame picture is overlapped, the accumulative overlapping number of each hot spot region frame is counted;
The overlapping density of each hot spot region frame is determined according to the accumulative overlapping number.
3. cheating method for detecting area as claimed in claim 2, which is characterized in that described to be overlapped to each frame picture, system
The accumulative overlapping number for counting each hot spot region frame, specifically includes:
Continuous each frame picture is successively overlapped, so that each hot spot region frame in each frame picture is merged;
The regional ensemble for obtaining fused each hot spot region frame obtains the accumulative of each hot spot region frame from the regional ensemble
It is overlapped number, the regional ensemble is obtained by following formula:
Wherein, n is iteration threshold, i.e. overlapping number between frame and frame, and i is current frame number, HiFor the hot spot region of present frame
Frame, W are regional ensemble.
4. cheating method for detecting area as claimed in claim 3, which is characterized in that each frame for obtaining video to be identified is drawn
The pixel coordinate position of person head in face generates number of people coordinate data according to the pixel coordinate position, specifically includes:
The video to be identified is identified, determines the person head target in each frame picture of the video to be identified;
The corresponding pixel coordinate position of person head target and person head quantity in each frame picture are obtained, by each pixel coordinate position
It sets with person head quantity as number of people coordinate data.
5. cheating method for detecting area as claimed in claim 4, which is characterized in that described to be obtained according to the number of people coordinate data
The hot spot region frame for obtaining each frame picture, specifically includes:
The corresponding target person headers box in each target person head is determined according to the number of people coordinate data;
Obtain the spacing in each frame picture between each target person headers box;
The spacing is less than the target person headers box of pre-determined distance threshold value as hot spot region frame.
6. cheating method for detecting area as claimed in claim 5, which is characterized in that described true according to the number of people coordinate data
Before determining the corresponding target person headers box in each target person head, the cheating method for detecting area further include:
From starting number of people coordinate data and the last frame picture extracted in the number of people coordinate data in first frame picture
Final number of people coordinate data;
The starting number of people coordinate data is compared with the final number of people coordinate data, obtains person head displacement ginseng
Number, the person head displacement parameter are the history displacement data of the pixel coordinate position on everyone head in first frame picture;
It is target according to the person head that the person head displacement parameter determines that person head displacement is less than preset displacement threshold value
Person head.
7. such as cheating method for detecting area of any of claims 1-6, which is characterized in that described to be overlapped density most
After the corresponding region of big hot spot region frame is as cheating region, the cheating method for detecting area further include:
The corresponding destination image data in the cheating region is obtained, the destination image data is substituting to default erroneous detection model
In, generate erroneous detection result;
It is, using the cheating region as prewarning area, to be generated according to the prewarning area pre- without mistaking in the erroneous detection result
Alert information, and the warning information is sent to monitoring center.
8. a kind of cheating regional detection device, which is characterized in that described device includes: data acquisition module, hot spot region determination
Module, density obtain module and cheating area determination module;
Wherein, the data acquisition module, the pixel coordinate position of person head in each frame picture for obtaining video to be identified
It sets, number of people coordinate data is generated according to the pixel coordinate position;
The hot spot region determining module, for obtaining the hot spot region frame of each frame picture according to the number of people coordinate data;
The density obtains module, and for each hot spot region frame to be iterated cluster, the overlapping for obtaining each hot spot region frame is close
Degree;
The cheating area determination module, for the corresponding region of the maximum hot spot region frame of density will to be overlapped as cheating area
Domain.
9. a kind of cheating equipment for area detection equipment, which is characterized in that the cheating equipment for area detection equipment includes: memory, processor
And it is stored in the cheating region detector that can be run on the memory and on the processor, the cheating region detection
Program is arranged for carrying out the step of cheating method for detecting area as described in any one of claims 1 to 7.
10. a kind of storage medium, which is characterized in that be stored with cheating region detector, the cheating on the storage medium
Realizing the cheating method for detecting area as described in any one of claims 1 to 7 when region detector is executed by processor
Step.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910042318.9A CN109815872A (en) | 2019-01-16 | 2019-01-16 | Cheating method for detecting area, device, equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910042318.9A CN109815872A (en) | 2019-01-16 | 2019-01-16 | Cheating method for detecting area, device, equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109815872A true CN109815872A (en) | 2019-05-28 |
Family
ID=66604471
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910042318.9A Pending CN109815872A (en) | 2019-01-16 | 2019-01-16 | Cheating method for detecting area, device, equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109815872A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111586369A (en) * | 2020-06-05 | 2020-08-25 | 上海商汤智能科技有限公司 | Aggregation detection method and device, electronic equipment and readable storage medium |
CN113701754A (en) * | 2021-09-06 | 2021-11-26 | 中国矿业大学(北京) | Underground three-dimensional accurate positioning system |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101916365A (en) * | 2010-07-06 | 2010-12-15 | 北京竞业达数码科技有限公司 | Intelligent video identifying method for cheat in test |
CN102799893A (en) * | 2012-06-15 | 2012-11-28 | 北京理工大学 | Method for processing monitoring video in examination room |
US20150193651A1 (en) * | 2014-01-03 | 2015-07-09 | Gleim Conferencing, Llc | System and method for validating test takers |
CN106713856A (en) * | 2016-12-15 | 2017-05-24 | 重庆凯泽科技股份有限公司 | Intelligent examination monitoring system and method |
CN107256225A (en) * | 2017-04-28 | 2017-10-17 | 济南中维世纪科技有限公司 | A kind of temperature drawing generating method and device based on video analysis |
CN107491717A (en) * | 2016-06-13 | 2017-12-19 | 科大讯飞股份有限公司 | The detection method that cheats at one's exam and device |
CN108304794A (en) * | 2018-01-26 | 2018-07-20 | 安徽爱学堂教育科技有限公司 | cheating automatic identifying method and device |
CN109034074A (en) * | 2018-07-31 | 2018-12-18 | 安徽皖新金智教育科技有限公司 | campus examination feedback system and feedback method |
-
2019
- 2019-01-16 CN CN201910042318.9A patent/CN109815872A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101916365A (en) * | 2010-07-06 | 2010-12-15 | 北京竞业达数码科技有限公司 | Intelligent video identifying method for cheat in test |
CN102799893A (en) * | 2012-06-15 | 2012-11-28 | 北京理工大学 | Method for processing monitoring video in examination room |
US20150193651A1 (en) * | 2014-01-03 | 2015-07-09 | Gleim Conferencing, Llc | System and method for validating test takers |
CN107491717A (en) * | 2016-06-13 | 2017-12-19 | 科大讯飞股份有限公司 | The detection method that cheats at one's exam and device |
CN106713856A (en) * | 2016-12-15 | 2017-05-24 | 重庆凯泽科技股份有限公司 | Intelligent examination monitoring system and method |
CN107256225A (en) * | 2017-04-28 | 2017-10-17 | 济南中维世纪科技有限公司 | A kind of temperature drawing generating method and device based on video analysis |
CN108304794A (en) * | 2018-01-26 | 2018-07-20 | 安徽爱学堂教育科技有限公司 | cheating automatic identifying method and device |
CN109034074A (en) * | 2018-07-31 | 2018-12-18 | 安徽皖新金智教育科技有限公司 | campus examination feedback system and feedback method |
Non-Patent Citations (1)
Title |
---|
吴凯: "头部运动分析及其在考场视频监控系统中的研究与应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111586369A (en) * | 2020-06-05 | 2020-08-25 | 上海商汤智能科技有限公司 | Aggregation detection method and device, electronic equipment and readable storage medium |
CN111586369B (en) * | 2020-06-05 | 2022-04-01 | 上海商汤智能科技有限公司 | Aggregation detection method and device, electronic equipment and readable storage medium |
CN113701754A (en) * | 2021-09-06 | 2021-11-26 | 中国矿业大学(北京) | Underground three-dimensional accurate positioning system |
CN113701754B (en) * | 2021-09-06 | 2023-05-12 | 中国矿业大学(北京) | Underground three-dimensional accurate positioning system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109670441B (en) | Method, system, terminal and computer readable storage medium for realizing wearing recognition of safety helmet | |
WO2019096174A1 (en) | Method and apparatus for garment color identification of figure images, and electronic device | |
WO2016114134A1 (en) | Motion condition estimation device, motion condition estimation method and program recording medium | |
JP6397581B2 (en) | Congestion status visualization device, congestion status visualization system, congestion status visualization method, and congestion status visualization program | |
US10846537B2 (en) | Information processing device, determination device, notification system, information transmission method, and program | |
CN106559695A (en) | Barrage message processing method and device and electronic equipment | |
CN105631418A (en) | People counting method and device | |
US9177214B1 (en) | Method and apparatus for an adaptive threshold based object detection | |
CN111767888A (en) | Object state detection method, computer device, storage medium, and electronic device | |
CN112464797B (en) | Smoking behavior detection method and device, storage medium and electronic equipment | |
WO2017092269A1 (en) | Passenger flow information collection method and apparatus, and passenger flow information processing method and apparatus | |
CN109508576A (en) | A kind of abnormal driving behavioral value method, apparatus and electronic equipment | |
CN109815872A (en) | Cheating method for detecting area, device, equipment and storage medium | |
CN107517391A (en) | A kind of method and apparatus for being used to differentiate abnormal live information in net cast | |
WO2022041484A1 (en) | Human body fall detection method, apparatus and device, and storage medium | |
CN111666915A (en) | Monitoring method, device, equipment and storage medium | |
CN113963316A (en) | Target event determination method and device, storage medium and electronic device | |
CN112464765B (en) | Safety helmet detection method based on single-pixel characteristic amplification and application thereof | |
CN105632003B (en) | The real-time estimating method and device of a kind of clearance port queuing time | |
CN109815902A (en) | A kind of pedestrian attribute region information acquisition method, device and equipment | |
CN105427303B (en) | A kind of vision measurement and method of estimation of substation's legacy | |
CN113947795B (en) | Mask wearing detection method, device, equipment and storage medium | |
WO2021103727A1 (en) | Data processing method, device, and storage medium | |
CN114387648A (en) | Abnormal image detection method, abnormal image detection device, computer equipment and storage medium | |
Ibrahim et al. | Mouth covered detection for yawn |
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 | ||
AD01 | Patent right deemed abandoned | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20210903 |