CN107945458A - Fall detection method, device and computing device under camera scene - Google Patents
Fall detection method, device and computing device under camera scene Download PDFInfo
- Publication number
- CN107945458A CN107945458A CN201711333987.9A CN201711333987A CN107945458A CN 107945458 A CN107945458 A CN 107945458A CN 201711333987 A CN201711333987 A CN 201711333987A CN 107945458 A CN107945458 A CN 107945458A
- Authority
- CN
- China
- Prior art keywords
- gesture recognition
- recognition result
- humanoid
- monitored object
- posture
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
- G08B21/0407—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis
- G08B21/043—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis detecting an emergency event, e.g. a fall
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Psychology (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Psychiatry (AREA)
- Multimedia (AREA)
- Social Psychology (AREA)
- General Health & Medical Sciences (AREA)
- Gerontology & Geriatric Medicine (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses fall detection method, device and the computing device under a kind of camera scene, wherein, method includes：Monitoring image under camera scene is obtained with prefixed time interval；Dividing processing is carried out to monitoring image, obtains humanoid figure's picture in monitoring image；By humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained；And determine whether humanoid figure falls as corresponding monitored object according to gesture recognition result；If so, then alarm signal is sent to monitor terminal.As can be seen here, utilize scheme provided by the invention, the gesture recognition module that training can be utilized to obtain, automatically detect whether monitored object falls under camera scene, and when monitored object is fallen, alarm signal is sent to monitor terminal in time, which dresses any equipment without monitored object, any operation is carried out without monitored object, therefore improves the automatic degree and convenience of fall detection.
Description
Technical field
The present invention relates to field of computer technology, and in particular to fall detection method, device under a kind of camera scene
And computing device.
Background technology
The fall detection of human body belongs to a part and parcel of human motion Activity recognition, in particular for being in alone
Old man, in the event of falling and nobody knows, it is possible to bring hardly imaginable consequence.Therefore, with social senilization
The acceleration of process, for " Empty nest elderly " accidentally tumble detection and remote emergency relief just by there is an urgent need to.
At present, for avoid old man because of accidentally tumble nobody know and caused by succour not in time the problems such as generation, can borrow
The wearable detection device of some detection user movement behaviors is helped, to detect the behavioral parameters during user movement, Jin Ershi
The detection now whether fallen to user.However, this wearable detection device needs user to carry with, and need to use
Family performs the operation such as unlatching, charging, it is impossible to which whether detection user falls in time in the state of no user operation.
The content of the invention
In view of the above problems, it is proposed that the present invention overcomes the above problem in order to provide one kind or solves at least in part
State fall detection method, device and the computing device under the camera scene of problem.
According to an aspect of the invention, there is provided the fall detection method under a kind of camera scene, including：
Monitoring image under camera scene is obtained with prefixed time interval；
Dividing processing is carried out to monitoring image, obtains humanoid figure's picture in monitoring image；
By humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained；And according to
Gesture recognition result determines whether humanoid figure falls as corresponding monitored object；
If so, then alarm signal is sent to monitor terminal.
Further, wherein, gesture recognition module trains to obtain by following steps：
Humanoid figure's picture of normal attitude and the humanoid figure of tumble posture are gathered as being used as sample image, sample image is carried out
Posture is demarcated；
Sample image is inputted into neutral net, neutral net is trained according to posture calibration result to obtain posture
Identification module.
Further, wherein, determine whether humanoid figure falls as corresponding monitored object according to gesture recognition result
Further comprise：
If gesture recognition result is normal attitude, it is determined that monitored object is not fallen；
If gesture recognition result is tumble posture, it is determined that monitored object is fallen.
Further, wherein, method further includes：
If gesture recognition result is tumble posture, interception includes the monitor video in the preset time period of monitoring image；
Monitor video is sent to monitor terminal.
Further, wherein, method further includes：
If gesture recognition result is tumble posture, the order of collected sound signal is sent to camera；
Crying detection is carried out to voice signal；
Sending alarm signal to monitor terminal is specially：If voice signal includes crying, send and alarm to monitor terminal
Signal.
Further, wherein, after gesture recognition result is obtained, method further includes：Whether judge gesture recognition result
For error detection result；
Determine humanoid figure is specially as whether corresponding monitored object occurs tumble according to gesture recognition result：If judge
Gesture recognition result is not error detection as a result, then determining whether humanoid figure sends out as corresponding monitored object according to gesture recognition result
It is raw to fall.
Further, wherein, judge whether gesture recognition result is that error detection result further comprises：
By manually marking whether definite gesture recognition result is error detection result；
Alternatively, the posture feature by the error detection marked in advance in posture feature and gesture recognition result and database
And gesture recognition result is compared, determine whether gesture recognition result is error detection result according to comparison result.
Further, wherein, by manually marking whether definite gesture recognition result is method after error detection result
Further include：
If by manually mark definite gesture recognition result as error detection as a result, if by posture feature and gesture recognition knot
The annotation results of fruit are stored into database.
According to another aspect of the present invention, there is provided the falling detection device under a kind of camera scene, including：
Acquisition module, suitable for obtaining the monitoring image under camera scene with prefixed time interval；
Split module, suitable for carrying out dividing processing to monitoring image, obtain humanoid figure's picture in monitoring image；
First detection module, suitable for as inputting into trained obtained gesture recognition module, humanoid figure is obtained posture
Recognition result；And determine whether humanoid figure falls as corresponding monitored object according to gesture recognition result；
Alarm module, if detecting that monitored object is fallen suitable for first detection module, sends to monitor terminal and reports
Alert signal.
Further, wherein, device further includes：Training module, is suitable for：Gather the humanoid figure's picture and tumble appearance of normal attitude
The humanoid figure of state carries out posture calibration as being used as sample image, to sample image；Sample image is inputted into neutral net, root
Neutral net is trained according to posture calibration result to obtain gesture recognition module.
Further, wherein, first detection module is further adapted for：
If gesture recognition result is normal attitude, it is determined that monitored object is not fallen；
If gesture recognition result is tumble posture, it is determined that monitored object is fallen.
Further, wherein, device further includes：
Interception module, if being tumble posture suitable for gesture recognition result, interception includes the preset time period of monitoring image
Interior monitor video；
Sending module, suitable for monitor video is sent to monitor terminal.
Further, wherein, device further includes：
Control module, if being tumble posture suitable for gesture recognition result, the life of collected sound signal is sent to camera
Order；
Second detection module, suitable for carrying out crying detection to voice signal；
Alarm module is further adapted for：If voice signal includes crying, alarm signal is sent to monitor terminal.
Further, wherein, device further includes：Judgment module, suitable for judging whether gesture recognition result is error detection knot
Fruit；
First detection module is further adapted for：If judge gesture recognition result be not error detection as a result, if according to posture
Recognition result determines whether humanoid figure falls as corresponding monitored object.
Further, wherein, judgment module is further adapted for：
By manually marking whether definite gesture recognition result is error detection result；
Alternatively, the posture feature by the error detection marked in advance in posture feature and gesture recognition result and database
And gesture recognition result is compared, determine whether gesture recognition result is error detection result according to comparison result.
Further, wherein, device further includes：Memory module, if suitable for by manually marking definite gesture recognition result
For error detection as a result, then by the storage of the annotation results of posture feature and gesture recognition result into database.
According to another aspect of the invention, there is provided a kind of computing device, including：Processor, memory, communication interface and
Communication bus, processor, memory and communication interface complete mutual communication by communication bus；
Memory is used to store an at least executable instruction, and executable instruction makes processor perform under above-mentioned camera scene
The corresponding operation of fall detection method.
In accordance with a further aspect of the present invention, there is provided a kind of computer-readable storage medium, is stored with least one in storage medium
Executable instruction, executable instruction make processor perform such as the corresponding operation of fall detection method under above-mentioned camera scene.
Fall detection method, device and computing device under camera scene according to the present invention, with prefixed time interval
Obtain the monitoring image under camera scene；Image segmentation is carried out to monitoring image, and is therefrom extracted with humanoid feature
Region is as humanoid figure's picture；By humanoid figure as inputting to gesture recognition model, identify humanoid figure as corresponding monitored object
Posture；Determine whether monitored object falls according to the posture of monitored object；If monitored object is fallen, to monitoring eventually
End sends alarm signal.It can be seen from the above that using scheme provided by the invention, the gesture recognition module that training can be utilized to obtain,
Automatically detect whether monitored object falls under camera scene, and when monitored object is fallen, in time to prison
Control terminal sends alarm signal, which dresses any equipment without monitored object, and any operation is carried out without monitored object,
Therefore the automatic degree and convenience of fall detection are improved.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention,
And can be practiced according to the content of specification, and in order to allow above and other objects of the present invention, feature and advantage can
Become apparent, below especially exemplified by the embodiment of the present invention.
Brief description of the drawings
By reading the detailed description of hereafter preferred embodiment, it is various other the advantages of and benefit it is common for this area
Technical staff will be clear understanding.Attached drawing is only used for showing the purpose of preferred embodiment, and is not considered as to the present invention
Limitation.And in whole attached drawing, identical component is denoted by the same reference numerals.In the accompanying drawings：
Fig. 1 shows the flow chart of the fall detection method under camera scene according to an embodiment of the invention；
Fig. 2 shows the flow chart of the fall detection method under camera scene in accordance with another embodiment of the present invention；
Fig. 3 shows the functional block diagram of the falling detection device under camera scene according to an embodiment of the invention；
Fig. 4 shows the functional block of the falling detection device under camera scene in accordance with another embodiment of the present invention
Figure；
Fig. 5 shows a kind of structure diagram of computing device according to embodiments of the present invention.
Embodiment
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although the disclosure is shown in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
Limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
Completely it is communicated to those skilled in the art.
Fig. 1 shows the flow chart of the fall detection method under camera scene according to an embodiment of the invention.Such as
Shown in Fig. 1, this method comprises the following steps：
Step S101：Monitoring image under camera scene is obtained with prefixed time interval.
Camera can monitor the scene under camera overlay area in real time, and be exported in the form of monitor video, and
The monitored results at each moment correspond to the frame monitoring image in monitor video.In this step, with prefixed time interval
The monitoring image at corresponding a certain moment is obtained, and then the monitored object that the monitoring image can be utilized to detect under camera scene is
It is no to fall.
Step S102：Dividing processing is carried out to monitoring image, obtains humanoid figure's picture in monitoring image.
The present invention carries out fall detection primarily directed to human body, therefore, it is necessary to be partitioned into from monitoring image in this step
Humanoid figure's picture.
Specifically, monitoring image is divided into according to features such as gray scale, color, texture, edge and shapes and some do not handed over mutually
Region repeatedly, and these features is showed similitude in the same area, and obvious difference is showed between different zones
Property；Then the region with humanoid feature is extracted from several regions being partitioned into.Wherein, monitoring image is split
The method of processing includes：Dividing method based on threshold value, the dividing method based on edge, the dividing method based on region, be based on
The dividing method of graph theory and/or the dividing method based on energy functional.
Step S103：By humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained；
And determine whether humanoid figure falls as corresponding monitored object according to gesture recognition result.
When carrying out fall detection, humanoid figure is as that the posture of corresponding monitored object has is normal and fall two kinds, and at this
In step, whether the gesture recognition module obtained by training determines monitored object to identify the posture of monitored object with this
Fall.
Specifically, first by humanoid figure as input is to trained obtained gesture recognition module, obtain gesture recognition as a result,
For example, gesture recognition result is normal or tumble.Wherein, gesture recognition module can be according to humanoid figure as reflecting and monitoring
The relevant feature of posture of object identifies the posture of monitored object.Then, determine that monitored object is according to gesture recognition result
No to fall, if monitored object is not fallen, this method terminates；If monitored object is fallen, to monitor terminal
Send alarm signal.
Step S104：Alarm signal is sent to monitor terminal.
If it is determined that monitored object is fallen, then in order to allow other people to know that someone falls in time, in order to
Corresponding urgent measure is taken for the situation.In this step, after definite monitored object is fallen, sent to monitor terminal
Alarm signal.
According to the fall detection method under camera scene provided in this embodiment, camera is obtained with prefixed time interval
Monitoring image under scene；For ease of carrying out fall detection for human body, image segmentation is carried out to monitoring image, and therefrom extract
The region of humanoid feature is provided as humanoid figure's picture；By humanoid figure as inputting to gesture recognition model, humanoid figure's picture is identified
The posture of corresponding monitored object；Determine whether monitored object falls according to the posture of monitored object；If monitored object is sent out
It is raw to fall, then alarm signal is sent to monitor terminal, so that other people can know that someone falls in time.It can be seen from the above that
Using scheme provided in this embodiment, the gesture recognition module that training can be utilized to obtain, is detected automatically under camera scene
Whether fall to monitored object, and when monitored object is fallen, send alarm signal, the mistake to monitor terminal in time
Cheng Wuxu monitored object dresses any equipment, carries out any operation without monitored object, therefore improve oneself of fall detection
Traverse degree and convenience.
Fig. 2 shows the flow chart of the fall detection method under camera scene in accordance with another embodiment of the present invention.
As shown in Fig. 2, this method comprises the following steps：
Step S201：Humanoid figure's picture of normal attitude and the humanoid figure of tumble posture are gathered as being used as sample image, to sample
This image carries out posture calibration；Sample image is inputted into neutral net, neutral net is carried out according to posture calibration result
Training obtains gesture recognition module.
In the present invention, mainly using gesture recognition module to humanoid figure as corresponding monitored object carry out posture knowledge
Not, and gesture recognition module then by substantial amounts of humanoid figure as being trained to obtain.In this step, substantial amounts of people is gathered
For shape image as sample image, the feature of the humanoid subject in sample image carries out posture calibration to each sample image；
Training input data using sample image as neutral net, the result of posture calibration export number as the training of neutral net
According to, adaptive learning is then carried out according to the training input data and training output data, it is trained to obtain gesture recognition module.
Specifically, humanoid figure is gathered from the monitoring video flow of monitoring device as being used as sample image, wherein, humanoid figure's picture
It can be obtained by humanoid detection technique or image Segmentation Technology from each frame monitoring image.Then, to humanoid in sample image
Object carry out posture calibration, generally, can by manually marking, it is semi-artificial mark or machine mark method carry out posture mark
Note.Optionally, in order to reduce the cost of mark, carried out by semi-artificial method, i.e.,：First with manually to a part of sample graph
As carrying out posture mark, for example, normal attitude is labeled as 1, tumble posture is labeled as 0；Then basis has marked out posture
Sample image feature, to remaining sample image carry out characteristic matching, sample image is gone out according to matching result automatic marking
The posture of corresponding humanoid subject.Sample image is inputted to the neutral net for setting initial parameter and inputs number as training
According to using the corresponding posture of each sample image as training output data, according to the training input data and training output data
Neutral net is trained, until error meets predetermined threshold value, then training is completed, and the neutral net that training obtains is appearance
State identification module.
Step S202：Monitoring image under camera scene is obtained with prefixed time interval.
Fall detection under camera scene, it is necessary first to obtain the monitoring image under camera scene.And due to people's
Tumble process, corresponds to the multiframe picture in monitor video, at this time, only a wherein frame monitoring image need to be detected, you can complete
The fall detection of paired monitored object.Therefore, in order to reduce the processing data amount of fall detection process, the efficiency of detection is improved,
In this step, the monitoring image under camera scene is obtained with prefixed time interval.Wherein, prefixed time interval can be according to detection
Demand flexibly determines, for example, if desired fall detection is more in time or more accurate, can reduce prefixed time interval.
Step S203：Dividing processing is carried out to monitoring image, obtains humanoid figure's picture in monitoring image.
In this step, humanoid figure's picture is partitioned into from monitoring image, in order to carry out gesture recognition for the humanoid image,
Influence of the feature in the non-humanoid region in monitoring image to gesture recognition process is avoided at the same time.Specifically, can directly utilize
Image Segmentation Technology is partitioned into humanoid figure's picture from monitoring image；But due to the complexity of monitoring image, add image point
The difficulty cut, in the present embodiment, can also first pass through humanoid detection technique and detect the corresponding monitored object institute of monitoring image
Region, be then partitioned into humanoid figure's picture using image Segmentation Technology in the region.
Step S204：By humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained.
In this step, by humanoid figure as input carries out gesture recognition to gesture recognition module, identify humanoid figure as
Posture is normal attitude or tumble posture.
Step S205：If gesture recognition result is tumble posture, the order of collected sound signal is sent to camera；It is right
Voice signal carries out crying detection.
In order to avoid gesture recognition result mistake cause false alarm the occurrence of, identifying humanoid figure as corresponding
The posture of monitored object is improved and fallen after tumble posture, to confirm the correctness of gesture recognition result by crying detection
The accuracy of testing result.
Specifically, after gesture recognition result is tumble posture, the order of collected sound signal is sent to camera.Shooting
Head is after the order is received, the voice signal under acquisition camera scene, optionally, when collection receives default after order
Interior voice signal.Then the feature of voice signal is extracted, for example, the feature of sound channel, mould is detected using trained crying
The feature of the voice signal of extraction is identified in block, determines whether include crying in voice signal according to recognition result.If bag
Containing crying, then confirm that gesture recognition result is correct, humanoid figure is tumble posture as the posture of corresponding monitored object；If do not include
Crying, then need to be further confirmed that the correctness of gesture recognition result.
In the gesture recognition by above-mentioned steps S204, and/or, after the crying detection of step S205, for gesture recognition
As a result it is humanoid figure's picture of normal attitude, and/or, crying testing result is the humanoid figure not comprising crying as, it is necessary under
Step S206 is stated to further confirm that the correctness of gesture recognition result.
Step S206：Judge whether gesture recognition result is error detection result；If it is not, then perform step S207；If so, then
According to humanoid figure as the actual posture of corresponding monitored object, take appropriate measures.
Since the gesture recognition module that training obtains still has identification error, i.e.,：It can not think gesture recognition module
Can be with posture of the entirely accurate without misrecognition all people's shape image, in this step, wrong report caused by order to reduce identification mistake
The occurrence of alert or leak detection, after gesture recognition result is obtained, determine whether gesture recognition result is flase drop
Survey as a result, to improve the accuracy rate of fall detection.For example, recognition result is normal attitude, and actually the humanoid image corresponds to
The posture of monitored object be tumble posture, it will cause the situation generation of leak detection；If judging, the gesture recognition result is
Error detection is as a result, then can be according to humanoid figure as the actual posture of corresponding monitored object, i.e.,：Tumble posture, takes corresponding
Measure, and then avoid the occurrence of leak detection.
Wherein, judge whether gesture recognition result is that the mode of error detection result there are two kinds：
Mode one, by manually marking whether definite gesture recognition result is error detection result.Specifically, know by posture
After other result feeds back to monitor terminal, for example, in the corresponding humanoid figure of recognition result for sending tumble posture as arriving monitor terminal
Afterwards, by artificial judgment, whether the corresponding monitored object of the humanoid image falls.If the result of artificial judgment is with passing through appearance
The result that state identification module identifies is inconsistent, it is determined that gesture recognition result is error detection as a result, correspondingly, marking out the people
The gesture recognition result of shape image is error detection result.
Mode two, the posture of the error detection of posture feature and gesture recognition result with having been marked in advance in database is special
Sign and gesture recognition result are compared, and determine whether gesture recognition result is error detection result according to comparison result.Specifically
Ground, when identify humanoid figure as corresponding monitored object posture after, extraction humanoid figure as the posture of corresponding monitored object
Feature, the gesture recognition result of posture feature and the humanoid image is compared with the error detection result stored in database
It is right, specifically by the posture feature extracted and the humanoid figure of error detection that has marked as corresponding monitored object posture feature into
Row compares, and gesture recognition result gesture recognition result corresponding with same error detection result is compared, if posture is special
Sign and the comparison of gesture recognition result are consistent, then judge for current humanoid figure as the gesture recognition of corresponding monitored object
As a result it is error detection result.
In addition, for the ease of improving the accuracy of fall detection with reference to the error detection result manually marked out, can also
Aforesaid way one and mode two are used in combination, and then more comprehensively judge error detection result.Specifically, if by artificial
The definite gesture recognition result of mark is error detection as a result, then arriving the annotation results storage of posture feature and gesture recognition result
In database, then can manually mark in combination one as a result, simultaneously Land use systems two are judged.
Step S207：If judge gesture recognition result be not error detection as a result, if people determined according to gesture recognition result
Whether the corresponding monitored object of shape image falls.If so, then perform step S208 and step S209；If it is not, then this method
Terminate.
If judge gesture recognition result be not error detection as a result, if show that gesture recognition result can reflect monitoring pair
As if no fallen.Specifically, if gesture recognition result is normal attitude, it is determined that monitored object is not fallen；If
Gesture recognition result is tumble posture, it is determined that monitored object is fallen.Further, however, it is determined that monitored object falls
, then need to feed back the result to monitor terminal in time.I.e.：When definite monitored object is fallen, then step S208 is performed
And step S209.
In addition, in another embodiment of the present invention, when judging that gesture recognition result is error detection as a result, then needing
According to humanoid figure as the actual posture of corresponding monitored object, take appropriate measures.Wherein, error detection is corresponding with two kinds of feelings
Condition, situation one, gesture recognition result are tumble posture, and actual is normal attitude；Situation two, gesture recognition result are normal appearance
State, and actual is tumble posture.Therefore, when judge for error detection result when, corresponding situation one, then this method terminate；It is corresponding
Situation two, it is determined that humanoid figure is fallen as corresponding monitored object, therefore, in this case, it is also desirable to perform this reality
Apply the following step S208 and step S209 of example.
Step S208：Alarm signal is sent to monitor terminal.
If it is determined that monitored object is fallen, then in order to allow other people to know that someone falls in time, in order to
Corresponding urgent measure is taken for the situation.In this step, after definite monitored object is fallen, sent to monitor terminal
Alarm signal.Wherein it is determined that the mode that monitored object is fallen includes：Monitored object is determined according to gesture recognition result
Fall, determine that monitored object is fallen according to crying testing result, and/or, determine that monitored object is sent out according to error detection result
It is raw to fall.Optionally, monitored object is determined there is a situation where falling for according to crying testing result, send and report to monitor terminal
Warning signal is specially：If voice signal includes crying, alarm signal is sent to monitor terminal.
Step S209：If gesture recognition result is tumble posture, interception is included in the preset time period of monitoring image
Monitor video；Monitor video is sent to monitor terminal.
While alarm signal is sent to monitor terminal, monitor video is sent to monitor terminal, in order to pass through prison
Control terminal intuitively determines the state of monitored object, for example, determining whether monitored object occurs accidentally tumble, and/or determines prison
Control the order of severity that object is fallen.
Specifically, if gesture recognition result is tumble posture, crying testing result determines to include crying, and/or, according to
Error detection result determines that monitored object is fallen, then interception includes the monitor video in the preset time period of monitoring image.Can
Choosing, in order to completely understand monitored object state front and rear at the time of monitoring image corresponds to, monitoring image is determined
For the middle two field picture of the monitor video of interception.
In addition, after the state of monitored object is determined by monitor terminal, however, it is determined that monitored object is not fallen, then
Corresponding gesture recognition result is determined as error detection as a result, and by humanoid figure as corresponding prison by way of manually marking
Control the posture feature of object and gesture recognition result is stored into database to be used as comparison standard when judging error detection.
According to the fall detection method under camera scene provided in this embodiment, the monitoring under the camera scene of acquisition
Image；Dividing processing is carried out to monitoring image, obtains humanoid figure's picture；The gesture recognition module obtained using training is to humanoid figure's picture
Corresponding monitored object carries out gesture recognition；It is tumble appearance in gesture recognition result to determine the correctness of gesture recognition result
During state, crying monitoring is carried out to the voice signal of camera collection, and, judge whether gesture recognition result is error detection knot
Fruit；Then determine whether monitored object falls according to gesture recognition result, and then improve the accuracy of fall detection；
After determining that monitored object is fallen, alarm signal is sent to monitor terminal, and interception includes the monitor video of monitoring image
It is sent to monitor terminal.It can be seen from the above that using scheme provided in this embodiment, the gesture recognition mould that training can be utilized to obtain
Block, automatically detects whether monitored object falls under camera scene；Also, with reference to crying monitoring result, and/or,
Error detection judge as a result, further confirm that whether monitored object falls, and then improve the accurate of fall detection result
Property；In the case where definite monitored object is fallen, alarm signal is sent to monitor terminal in time and sends monitor video,
In order to intuitively determine the state of monitored object in time by monitor terminal.
Fig. 3 shows the functional block diagram of the falling detection device under camera scene according to an embodiment of the invention.
As shown in figure 3, the device includes：
Acquisition module 301, suitable for obtaining the monitoring image under camera scene with prefixed time interval；
Split module 302, suitable for carrying out dividing processing to monitoring image, obtain humanoid figure's picture in monitoring image；
First detection module 303, suitable for as inputting into trained obtained gesture recognition module, humanoid figure is obtained appearance
State recognition result；And determine whether humanoid figure falls as corresponding monitored object according to gesture recognition result；
Alarm module 304, if detecting that monitored object is fallen suitable for first detection module, sends to monitor terminal
Alarm signal.
Fig. 4 shows the functional block of the falling detection device under camera scene in accordance with another embodiment of the present invention
Figure.As shown in figure 4, on the basis of Fig. 3, which further includes：Training module 401, interception module 402, sending module 403,
Control module 404, the second detection module 405, judgment module 406 and memory module 407.
Training module 401, is suitable for：Humanoid figure's picture of normal attitude and the humanoid figure of tumble posture are gathered as being used as sample graph
Picture, posture calibration is carried out to sample image；Sample image is inputted into neutral net, according to posture calibration result to nerve net
Network is trained to obtain gesture recognition module.
First detection module 303 is further adapted for：
If gesture recognition result is normal attitude, it is determined that monitored object is not fallen；
If gesture recognition result is tumble posture, it is determined that monitored object is fallen.
Interception module 402, if being tumble posture suitable for gesture recognition result, interception includes the preset time of monitoring image
Monitor video in section；
Sending module 403, suitable for monitor video is sent to monitor terminal.
Control module 404, if being tumble posture suitable for gesture recognition result, collected sound signal is sent to camera
Order；
Second detection module 405, suitable for carrying out crying detection to voice signal；
Alarm module 304 is further adapted for：If voice signal includes crying, alarm signal is sent to monitor terminal.
Judgment module 406, suitable for judging whether gesture recognition result is error detection result；
First detection module 303 is further adapted for：If judge gesture recognition result be not error detection as a result, if according to appearance
State recognition result determines whether humanoid figure falls as corresponding monitored object.
Judgment module 406 is further adapted for：
By manually marking whether definite gesture recognition result is error detection result；
Alternatively, the posture feature by the error detection marked in advance in posture feature and gesture recognition result and database
And gesture recognition result is compared, determine whether gesture recognition result is error detection result according to comparison result.
Memory module 407, suitable for if by manually mark definite gesture recognition result as error detection as a result, if posture is special
The annotation results of sign and gesture recognition result are stored into database.
Concrete structure and operation principle on above-mentioned modules can refer to the description of corresponding steps in embodiment of the method,
Details are not described herein again.
The embodiment of the present application provides a kind of nonvolatile computer storage media, and computer-readable storage medium is stored with least
One executable instruction, the computer executable instructions can perform the tumble under the camera scene in above-mentioned any means embodiment
Detection method.
Fig. 5 shows a kind of structure diagram of computing device according to embodiments of the present invention, the specific embodiment of the invention
The specific implementation to computing device does not limit.
As shown in figure 5, the computing device can include：Processor (processor) 502, communication interface
(Communications Interface) 504, memory (memory) 506 and communication bus 508.
Wherein：
Processor 502, communication interface 504 and memory 506 complete mutual communication by communication bus 508.
Communication interface 504, for communicating with the network element of miscellaneous equipment such as client or other servers etc..
Processor 502, for executive program 510, can specifically perform the correlation step in above method embodiment.
Specifically, program 510 can include program code, which includes computer-managed instruction.
Processor 502 is probably central processor CPU, or specific integrated circuit ASIC (Application
Specific Integrated Circuit), or be arranged to implement the embodiment of the present invention one or more integrate electricity
Road.The one or more processors that computing device includes, can be same type of processors, such as one or more CPU；Also may be used
To be different types of processor, such as one or more CPU and one or more ASIC.
Memory 506, for storing program 510.Memory 506 may include high-speed RAM memory, it is also possible to further include
Nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.
Program 510 specifically can be used for so that processor 502 performs following operation：
Monitoring image under camera scene is obtained with prefixed time interval；
Dividing processing is carried out to monitoring image, obtains humanoid figure's picture in monitoring image；
By humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained；And according to
Gesture recognition result determines whether humanoid figure falls as corresponding monitored object；
If so, then alarm signal is sent to monitor terminal.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
Humanoid figure's picture of normal attitude and the humanoid figure of tumble posture are gathered as being used as sample image, sample image is carried out
Posture is demarcated；
Sample image is inputted into neutral net, neutral net is trained according to posture calibration result to obtain posture
Identification module.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
If gesture recognition result is normal attitude, it is determined that monitored object is not fallen；
If gesture recognition result is tumble posture, it is determined that monitored object is fallen.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
If gesture recognition result is tumble posture, interception includes the monitor video in the preset time period of monitoring image；
Monitor video is sent to monitor terminal.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
If gesture recognition result is tumble posture, the order of collected sound signal is sent to camera；
Crying detection is carried out to voice signal；
Sending alarm signal to monitor terminal is specially：If voice signal includes crying, send and alarm to monitor terminal
Signal.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：Judge whether gesture recognition result is error detection result；
Determine humanoid figure is specially as whether corresponding monitored object occurs tumble according to gesture recognition result：If judge
Gesture recognition result is not error detection as a result, then determining whether humanoid figure sends out as corresponding monitored object according to gesture recognition result
It is raw to fall.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
By manually marking whether definite gesture recognition result is error detection result；
Alternatively, the posture feature by the error detection marked in advance in posture feature and gesture recognition result and database
And gesture recognition result is compared, determine whether gesture recognition result is error detection result according to comparison result.
In a kind of optional mode, program 510 can specifically be further used for so that processor 502 performs following behaviour
Make：
If by manually mark definite gesture recognition result as error detection as a result, if by posture feature and gesture recognition knot
The annotation results of fruit are stored into database.
Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein.
Various general-purpose systems can also be used together with teaching based on this.As described above, required by constructing this kind of system
Structure be obvious.In addition, the present invention is not also directed to any certain programmed language.It should be understood that it can utilize various
Programming language realizes the content of invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the specification that this place provides, numerous specific details are set forth.It is to be appreciated, however, that the implementation of the present invention
Example can be put into practice in the case of these no details.In some instances, known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help to understand one or more of each inventive aspect,
Above in the description to the exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：I.e. required guarantor
The application claims of shield features more more than the feature being expressly recited in each claim.It is more precisely, such as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following embodiment are expressly incorporated in the embodiment, wherein each claim is in itself
Separate embodiments all as the present invention.
Those skilled in the art, which are appreciated that, to carry out adaptively the module in the equipment in embodiment
Change and they are arranged in one or more equipment different from the embodiment.Can be the module or list in embodiment
Member or component be combined into a module or unit or component, and can be divided into addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit exclude each other, it can use any
Combination is disclosed to all features disclosed in this specification (including adjoint claim, summary and attached drawing) and so to appoint
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (including adjoint power
Profit requires, summary and attached drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation
Replace.
In addition, it will be appreciated by those of skill in the art that although some embodiments in this include institute in other embodiments
Including some features rather than further feature, but the combination of the feature of different embodiment means to be in the scope of the present invention
Within and form different embodiments.For example, in the following claims, embodiment claimed it is any it
One mode can use in any combination.
The all parts embodiment of the present invention can be with hardware realization, or to be run on one or more processor
Software module realize, or realized with combinations thereof.It will be understood by those of skill in the art that it can use in practice
Microprocessor or digital signal processor (DSP) realize the fall detection under camera scene according to embodiments of the present invention
The some or all functions of some or all components in device.The present invention is also implemented as being used to perform being retouched here
The some or all equipment or program of device (for example, computer program and computer program product) for the method stated.
Such program for realizing the present invention can store on a computer-readable medium, or can have one or more signal
Form.Such signal can be downloaded from internet website and obtained, either provide on carrier signal or with it is any its
He provides form.
It should be noted that the present invention will be described rather than limits the invention for above-described embodiment, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of some different elements and being come by means of properly programmed computer real
It is existing.In if the unit claim of equipment for drying is listed, several in these devices can be by same hardware branch
To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame
Claim.
Claims (10)
1. the fall detection method under a kind of camera scene, including：
Monitoring image under camera scene is obtained with prefixed time interval；
Dividing processing is carried out to the monitoring image, obtains humanoid figure's picture in the monitoring image；
By the humanoid figure as inputting into trained obtained gesture recognition module, gesture recognition result is obtained；And according to
The gesture recognition result determines whether humanoid figure falls as corresponding monitored object；
If so, then alarm signal is sent to monitor terminal.
2. according to the method described in claim 1, wherein, the gesture recognition module trains to obtain by following steps：
Humanoid figure's picture of normal attitude and the humanoid figure of tumble posture are gathered as being used as sample image, posture is carried out to sample image
Calibration；
The sample image is inputted into neutral net, neutral net is trained according to posture calibration result to obtain posture
Identification module.
3. method according to claim 1 or 2, wherein, it is described to determine humanoid figure as corresponding according to gesture recognition result
Whether monitored object, which occurs tumble, further comprises：
If the gesture recognition result is normal attitude, it is determined that the monitored object is not fallen；
If the gesture recognition result is tumble posture, it is determined that the monitored object is fallen.
4. method according to any one of claim 1-3, wherein, the method further includes：
If the gesture recognition result is tumble posture, the monitoring that intercepting is included in the preset time period of the monitoring image regards
Frequently；
The monitor video is sent to the monitor terminal.
5. according to the described method of any one of claim 1-4, wherein, the method further includes：
If the gesture recognition result is tumble posture, the order of collected sound signal is sent to camera；
Crying detection is carried out to the voice signal；
It is described to be specially to monitor terminal transmission alarm signal：If the voice signal includes crying, sent to monitor terminal
Alarm signal.
6. according to the method any one of claim 1-5, wherein, after gesture recognition result is obtained, the method
Further include：Judge whether the gesture recognition result is error detection result；
It is described to determine humanoid figure is specially as whether corresponding monitored object occurs tumble according to gesture recognition result：If judge
The gesture recognition result is not error detection as a result, then determining that humanoid figure is according to gesture recognition result as corresponding monitored object
It is no to fall.
It is 7. described to judge whether gesture recognition result is that error detection result is further according to the method described in claim 6, wherein
Including：
By manually marking whether definite gesture recognition result is error detection result；
Alternatively, by posture feature and gesture recognition result and the posture feature of error detection that has been marked in advance in database and
Gesture recognition result is compared, and determines whether gesture recognition result is error detection result according to comparison result.
8. the falling detection device under a kind of camera scene, including：
Acquisition module, suitable for obtaining the monitoring image under camera scene with prefixed time interval；
Split module, suitable for carrying out dividing processing to the monitoring image, obtain humanoid figure's picture in the monitoring image；
First detection module, suitable for as inputting into trained obtained gesture recognition module, the humanoid figure is obtained posture
Recognition result；And determine whether humanoid figure falls as corresponding monitored object according to the gesture recognition result；
Alarm module, if detecting that the monitored object is fallen suitable for first detection module, sends to monitor terminal and reports
Alert signal.
9. a kind of computing device, including：Processor, memory, communication interface and communication bus, the processor, the storage
Device and the communication interface complete mutual communication by the communication bus；
The memory is used to store an at least executable instruction, and the executable instruction makes the processor perform right such as will
Ask the corresponding operation of fall detection method under the camera scene any one of 1-7.
10. a kind of computer-readable storage medium, an at least executable instruction, the executable instruction are stored with the storage medium
Processor is set to perform the corresponding operation of fall detection method under the camera scene as any one of claim 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711333987.9A CN107945458A (en) | 2017-12-11 | 2017-12-11 | Fall detection method, device and computing device under camera scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711333987.9A CN107945458A (en) | 2017-12-11 | 2017-12-11 | Fall detection method, device and computing device under camera scene |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107945458A true CN107945458A (en) | 2018-04-20 |
Family
ID=61943083
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711333987.9A Pending CN107945458A (en) | 2017-12-11 | 2017-12-11 | Fall detection method, device and computing device under camera scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107945458A (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108966196A (en) * | 2018-08-07 | 2018-12-07 | 河南云拓智能科技有限公司 | A kind of data monitoring method based on Bluetooth gateway |
| CN109003420A (en) * | 2018-08-07 | 2018-12-14 | 河南云拓智能科技有限公司 | A kind of data monitoring method based on Zigbee gateway |
| CN109051629A (en) * | 2018-07-02 | 2018-12-21 | 成都精位科技有限公司 | The determination method, apparatus and system that material is fallen |
| CN109543607A (en) * | 2018-11-22 | 2019-03-29 | 广州小狗机器人技术有限公司 | Object abnormal state detection method, system, monitor system and storage medium |
| CN110263634A (en) * | 2019-05-13 | 2019-09-20 | 平安科技（深圳）有限公司 | Monitoring method, device, computer equipment and the storage medium of monitoring objective |
| CN110599736A (en) * | 2019-09-02 | 2019-12-20 | 广东颐寿医疗养老有限公司 | Personnel nursing early warning control method and system |
| CN110602323A (en) * | 2019-09-12 | 2019-12-20 | 奇酷互联网络科技（深圳）有限公司 | Tumble monitoring method, terminal and storage device |
| CN110686353A (en) * | 2019-09-26 | 2020-01-14 | 珠海格力电器股份有限公司 | Image control method and device, storage medium and air conditioner |
| CN111127837A (en) * | 2018-10-31 | 2020-05-08 | 杭州海康威视数字技术股份有限公司 | Alarm method, camera and alarm system |
| CN111275928A (en) * | 2020-02-20 | 2020-06-12 | 北京澎思科技有限公司 | Early warning method, electronic equipment, terminal equipment and early warning system for monitoring human body activity |
| CN111904429A (en) * | 2020-07-30 | 2020-11-10 | 中国建设银行股份有限公司 | Human body falling detection method and device, electronic equipment and storage medium |
| CN112215185A (en) * | 2020-10-21 | 2021-01-12 | 成都信息工程大学 | System and method for detecting falling behavior from monitoring video |
| CN112669316A (en) * | 2021-01-29 | 2021-04-16 | 南方电网调峰调频发电有限公司 | Power production abnormity monitoring method and device, computer equipment and storage medium |
| CN112733618A (en) * | 2020-12-22 | 2021-04-30 | 江苏艾雨文承养老机器人有限公司 | Human body falling detection method, falling prevention robot and falling prevention system |
| CN112906548A (en) * | 2021-02-07 | 2021-06-04 | 广东省科学院智能制造研究所 | Fall detection method and system based on edge calculation |
| WO2021164654A1 (en) * | 2020-02-20 | 2021-08-26 | 艾科科技股份有限公司 | Time-continuity-based detection determination system and method |
| CN113497921A (en) * | 2021-07-07 | 2021-10-12 | 合肥英睿系统技术有限公司 | Monitoring method and intelligent monitoring system |
| CN115273395A (en) * | 2022-05-31 | 2022-11-01 | 歌尔股份有限公司 | Monitoring method, device, equipment, system and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105023022A (en) * | 2015-07-09 | 2015-11-04 | 深圳天珑无线科技有限公司 | Tumble detection method and system |
| CN105448041A (en) * | 2016-01-22 | 2016-03-30 | 苏州望湖房地产开发有限公司 | A human body falling intelligent control system and method |
| CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
-
2017
- 2017-12-11 CN CN201711333987.9A patent/CN107945458A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105023022A (en) * | 2015-07-09 | 2015-11-04 | 深圳天珑无线科技有限公司 | Tumble detection method and system |
| CN105448041A (en) * | 2016-01-22 | 2016-03-30 | 苏州望湖房地产开发有限公司 | A human body falling intelligent control system and method |
| CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109051629B (en) * | 2018-07-02 | 2020-09-04 | 成都精位科技有限公司 | Method, device and system for determining falling of materials |
| CN109051629A (en) * | 2018-07-02 | 2018-12-21 | 成都精位科技有限公司 | The determination method, apparatus and system that material is fallen |
| CN109003420A (en) * | 2018-08-07 | 2018-12-14 | 河南云拓智能科技有限公司 | A kind of data monitoring method based on Zigbee gateway |
| CN108966196A (en) * | 2018-08-07 | 2018-12-07 | 河南云拓智能科技有限公司 | A kind of data monitoring method based on Bluetooth gateway |
| CN108966196B (en) * | 2018-08-07 | 2020-09-08 | 河南云拓智能科技有限公司 | Data monitoring method based on Bluetooth gateway |
| CN111127837A (en) * | 2018-10-31 | 2020-05-08 | 杭州海康威视数字技术股份有限公司 | Alarm method, camera and alarm system |
| CN109543607A (en) * | 2018-11-22 | 2019-03-29 | 广州小狗机器人技术有限公司 | Object abnormal state detection method, system, monitor system and storage medium |
| CN110263634A (en) * | 2019-05-13 | 2019-09-20 | 平安科技（深圳）有限公司 | Monitoring method, device, computer equipment and the storage medium of monitoring objective |
| CN110599736A (en) * | 2019-09-02 | 2019-12-20 | 广东颐寿医疗养老有限公司 | Personnel nursing early warning control method and system |
| CN110599736B (en) * | 2019-09-02 | 2021-06-08 | 广东颐寿医疗养老有限公司 | Personnel nursing early warning control method and system |
| CN110602323A (en) * | 2019-09-12 | 2019-12-20 | 奇酷互联网络科技（深圳）有限公司 | Tumble monitoring method, terminal and storage device |
| CN110602323B (en) * | 2019-09-12 | 2021-04-20 | 奇酷互联网络科技（深圳）有限公司 | Tumble monitoring method, terminal and storage device |
| CN110686353A (en) * | 2019-09-26 | 2020-01-14 | 珠海格力电器股份有限公司 | Image control method and device, storage medium and air conditioner |
| WO2021164654A1 (en) * | 2020-02-20 | 2021-08-26 | 艾科科技股份有限公司 | Time-continuity-based detection determination system and method |
| CN111275928A (en) * | 2020-02-20 | 2020-06-12 | 北京澎思科技有限公司 | Early warning method, electronic equipment, terminal equipment and early warning system for monitoring human body activity |
| CN111904429A (en) * | 2020-07-30 | 2020-11-10 | 中国建设银行股份有限公司 | Human body falling detection method and device, electronic equipment and storage medium |
| CN112215185A (en) * | 2020-10-21 | 2021-01-12 | 成都信息工程大学 | System and method for detecting falling behavior from monitoring video |
| CN112733618A (en) * | 2020-12-22 | 2021-04-30 | 江苏艾雨文承养老机器人有限公司 | Human body falling detection method, falling prevention robot and falling prevention system |
| CN112669316A (en) * | 2021-01-29 | 2021-04-16 | 南方电网调峰调频发电有限公司 | Power production abnormity monitoring method and device, computer equipment and storage medium |
| CN112906548A (en) * | 2021-02-07 | 2021-06-04 | 广东省科学院智能制造研究所 | Fall detection method and system based on edge calculation |
| CN112906548B (en) * | 2021-02-07 | 2024-05-28 | 广东省科学院智能制造研究所 | Fall detection method and system based on edge calculation |
| CN113497921A (en) * | 2021-07-07 | 2021-10-12 | 合肥英睿系统技术有限公司 | Monitoring method and intelligent monitoring system |
| CN115273395A (en) * | 2022-05-31 | 2022-11-01 | 歌尔股份有限公司 | Monitoring method, device, equipment, system and storage medium |
| CN115273395B (en) * | 2022-05-31 | 2024-03-12 | 歌尔股份有限公司 | Monitoring method, device, equipment, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107945458A (en) | Fall detection method, device and computing device under camera scene | |
| CN110097037B (en) | Intelligent monitoring method and device, storage medium and electronic equipment | |
| US11176418B2 (en) | Model test methods and apparatuses | |
| CN107679504A (en) | Face identification method, device, equipment and storage medium based on camera scene | |
| CN108647652A (en) | A kind of cotton development stage automatic identifying method based on image classification and target detection | |
| CN109544862A (en) | Activity recognition method, apparatus, storage medium and equipment based on smart home | |
| CN104076809B (en) | Data processing equipment and data processing method | |
| CN108696486B (en) | Abnormal operation behavior detection processing method and device | |
| CN109492665A (en) | Detection method, device and the electronic equipment of growth period duration of rice | |
| Bahaghighat et al. | Vision inspection of bottle caps in drink factories using convolutional neural networks | |
| CN107895172A (en) | Utilize the method, apparatus and computing device of image information detection anomalous video file | |
| CN110411554B (en) | Motor equipment detection method, device and system | |
| CN108182444A (en) | The method and device of video quality diagnosis based on scene classification | |
| CN113591751A (en) | Transformer substation abnormal condition warning method and device, computer equipment and storage medium | |
| CN117435999A (en) | Risk assessment method, apparatus, device and medium | |
| CN107729873A (en) | Special object detection method, device and computing device in video file | |
| CN105553767B (en) | Website backdoor file detection method and device | |
| CN113837390A (en) | Modal information completion method, device and equipment | |
| CN105468887B (en) | Data analysis system and method | |
| CN110310260A (en) | Sub-material decision-making technique, equipment and storage medium based on machine learning model | |
| CN108875453A (en) | The method, apparatus and computer storage medium of face picture bottom library registration | |
| CN113837138A (en) | Dressing monitoring method, system, medium and electronic terminal | |
| CN108694357A (en) | Method, apparatus and computer storage media for In vivo detection | |
| CN108829888A (en) | A kind of non-relational database document management method and device | |
| CN112798955B (en) | Fault detection method and device for special motor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180420 |