CN109887234A - A kind of children loss prevention method, apparatus, electronic equipment and storage medium - Google Patents

A kind of children loss prevention method, apparatus, electronic equipment and storage medium Download PDF

Info

Publication number
CN109887234A
CN109887234A CN201910172200.8A CN201910172200A CN109887234A CN 109887234 A CN109887234 A CN 109887234A CN 201910172200 A CN201910172200 A CN 201910172200A CN 109887234 A CN109887234 A CN 109887234A
Authority
CN
China
Prior art keywords
children
original image
target scene
monitoring video
video flow
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
Application number
CN201910172200.8A
Other languages
Chinese (zh)
Inventor
杨尊程
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Baidu Online Network Technology Beijing Co Ltd
Original Assignee
Baidu Online Network Technology Beijing Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Baidu Online Network Technology Beijing Co Ltd filed Critical Baidu Online Network Technology Beijing Co Ltd
Priority to CN201910172200.8A priority Critical patent/CN109887234A/en
Publication of CN109887234A publication Critical patent/CN109887234A/en
Pending legal-status Critical Current

Links

Abstract

The invention discloses a kind of children loss prevention method, apparatus, electronic equipment and storage mediums, which comprises obtains the monitoring video flow of target scene;It whether is determined in the target scene based on the monitoring video flow with the presence of children;If it is determined that then carrying out depth analysis to the monitoring video flow with the presence of children in the target scene, analysis result is obtained;Determine the children with the presence or absence of risk of wandering away based on the analysis result;If the children have risk of wandering away, alarm.By using above-mentioned technical proposal, the purpose that all children in wide area are carried out with anti-monitoring of wandering away is realized.

Description

A kind of children loss prevention method, apparatus, electronic equipment and storage medium
Technical field
The present embodiments relate to protection of the child technology more particularly to a kind of children loss prevention method, apparatus, electronic equipment And storage medium.
Background technique
In the place that the crowd is dense such as megastore, business district, because the crowd is dense, the reasons such as crowded, it is easy to occur The case where children loss.
Currently, the scheme of the children loss prevention generallyd use are as follows: by the way that positioning device is worn with children, which is set It is ready for use on and the position of children is positioned, and carry out information exchange with the intelligent terminal of children guardian in real time, once inspection When measuring the distance between children and its guardian farther out, then alarm.
In the scheme of above-mentioned children loss prevention, need to purchase and wear the positioning device, cost in advance for every children It is higher, and when finding children loss, the information at children loss scene can not be retained in time, once children fall into offender's In hand, the positioning device that children wear can be dropped and destroy.And the crowd is dense for megastore, business district etc. Scene does not ensure that every children wear the positioning device, and above scheme is for no youngster for wearing the positioning device Tong Ze, which cannot achieve, prevents missing purpose.
Summary of the invention
The present invention provides a kind of children loss prevention method, apparatus, electronic equipment and storage medium, realizes to large area model All children in enclosing carry out the purpose of anti-monitoring of wandering away.
In a first aspect, the embodiment of the invention provides a kind of children loss prevention methods, this method comprises:
Obtain the monitoring video flow of target scene;
It whether is determined in the target scene based on the monitoring video flow with the presence of children;
If it is determined that then carrying out depth analysis with the presence of children in the target scene to the monitoring video flow, being divided Analyse result;
Determine the children with the presence or absence of risk of wandering away based on the analysis result;
If the children have risk of wandering away, alarm.
Second aspect, the embodiment of the invention also provides a kind of children loss prevention device, which includes:
Module is obtained, for obtaining the monitoring video flow of target scene;
Children's determining module, for whether being determined in the target scene based on the monitoring video flow with the presence of children;
Analysis module, for if it is determined that then being carried out to the monitoring video flow deep with the presence of children in the target scene Degree analysis obtains analysis result;
Risk determining module, for determining the children with the presence or absence of risk of wandering away based on the analysis result;
Alarm module is alarmed if there is risk of wandering away for the children.
The third aspect, the embodiment of the invention also provides a kind of electronic equipment, the electronic equipment includes:
One or more processors;
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processing Device realizes such as children loss prevention method of any of claims 1-10.
Fourth aspect, the embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer Program realizes such as children loss prevention method of any of claims 1-10 when the program is executed by processor.
A kind of children loss prevention method provided in an embodiment of the present invention, by the monitoring video flow for obtaining target scene;Base It whether determines in the target scene in the monitoring video flow with the presence of children;If it is determined that thering are children to deposit in the target scene Depth analysis then is being carried out to the monitoring video flow, is obtaining analysis result;Determine that the children are based on the analysis result It is no to there is risk of wandering away;If the children have risk of wandering away, the technological means alarmed is realized to wide area Interior all children carry out the purpose of anti-monitoring of wandering away.
Detailed description of the invention
Fig. 1 is the flow diagram of one of embodiment of the present invention one children loss prevention method;
Fig. 2 is the flow diagram of one of embodiment of the present invention two children loss prevention method;
Fig. 3 is the flow diagram of one of embodiment of the present invention three children loss prevention method;
Fig. 4 is the flow diagram of one of embodiment of the present invention four children loss prevention method;
Fig. 5 is the flow diagram of one of embodiment of the present invention five children loss prevention method;
Fig. 6 is the flow diagram of one of embodiment of the present invention six children loss prevention method;
Fig. 7 is the structural schematic diagram of one of the embodiment of the present invention seven children loss prevention device;
Fig. 8 is the structural schematic diagram of one of the embodiment of the present invention eight electronic equipment.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 is a kind of flow diagram for children loss prevention method that the embodiment of the present invention one provides, and the present embodiment can fit Security monitoring is carried out for the children to the public place that the crowd is dense such as market, business district, it, should the case where with children loss prevention Method can be executed by children loss prevention device, and wherein the device can be implemented by software and/or hardware.Referring specifically to Fig. 1 institute Show, the children loss prevention method includes the following steps:
Step 110, the monitoring video flow for obtaining target scene.
Wherein, the target scene includes the large area place such as megastore, business district, can also be any including supermarket etc. Need to carry out the place of children's safety monitoring.The monitoring video flow can be obtained by the camera being arranged in around target scene, The quantity of the camera can be set according to the area of target scene, and the purpose is to realize to supervise 360 degree of target scene without dead angle Control, to achieve the purpose that the children to position any in target scene carry out security monitoring.
Whether step 120 is determined in the target scene based on the monitoring video flow with the presence of children, however, it is determined that described With the presence of children in target scene, then step 130 is continued to execute.
Specifically, obtaining the continuous original image of at least two frames based on the monitoring video flow, it is then based on the original graph Piece combination face recognition technology identifies that all faces in original image identify and speculate then in conjunction with picture recognition technology The corresponding age information of each face, finally determines in target scene whether have children to deposit according to the corresponding age information of each face ?.Or using everyone clothing type in picture recognition technology identification original image, target field is determined by clothing type Whether with the presence of children in scape, the clothing type is specific can include: child clothing and adult dress ornament.It can also be known by face Everyone height data, determines target scene by height data in other technology and picture recognition technology identification original image In whether with the presence of children.
Wherein, whether determined in the target scene based on the monitoring video flow has existing for children operation can be by setting The monitor terminal set around target scene executes, and can also be executed by cloud Analysis server;If by cloud Analysis server The operation is executed, then the monitoring video flow of the target scene obtained in real time need to be uploaded in real time cloud Analysis Service by monitor terminal Device.
Step 130 carries out depth analysis to the monitoring video flow, obtains analysis result.
Wherein, the essence of depth analysis is carried out to the monitoring video flow are as follows: to the mood shape of children in monitoring video flow State is analyzed, such as the children in analysis monitoring video flow, either with or without crying and screaming, anxious or panic etc. can reveal is worked as There is the Emotion expression for risk of wandering away in preceding children.Specifically, depth learning technology can be based on, to the youngster showed with specific emotional Virgin monitoring video flow is learnt in advance, to remember the feature of various Emotion expressions, is realized to youngster in new monitoring video flow The purpose that virgin Emotion expression is identified.
Specifically, the above-mentioned operation for carrying out depth analysis to the monitoring video flow can be executed by cloud Analysis server, If the operation " whether determined in the target scene based on the monitoring video flow with the presence of children " in above-mentioned steps 120 is by setting The monitor terminal set around target scene executes, then in the presence of monitor terminal determines and has children in target scene, then needs Monitoring video flow including the children is sent to cloud Analysis server, so that cloud Analysis server is realized to the monitoring The purpose of video flowing progress depth analysis.
Step 140 determines the children based on the analysis result with the presence or absence of risk of wandering away, if the children exist It goes wrong danger, thens follow the steps 150.
Wherein, the children can be determined according to the emotional state of children with the presence or absence of risk of wandering away, for example, if current child Emotional state be it is anxious, cry and scream or panic, it is determined that there is risk of wandering away in current child.In order to whether improve children In the presence of the determination precision for risk of wandering away, can be combined on the basis of the emotional state according to children around children adult's information into Row determines, for example, if the emotional state of current child is anxiety, and when there is no adult around current child, it is determined that it is current There is risk of wandering away in children.Alternatively, and having adult around current child if the emotional state of current child is anxiety, still The sight or focus of the adult be not on current child, it is determined that current child has risk of wandering away.Alternatively, if current The emotional state of children is anxiety, and has adult around current child, but the adult and current child do not have blood relationship pass When being, it is determined that current child has risk of wandering away.
Further, the children loss prevention method further include: will include the youngster if the children have risk of wandering away The monitoring video flow of the virgin target scene is saved, and is sent to the corresponding terminal device of Security Personnel and (such as is monitored The computer of room), it to carry out manual review, improves monitored children loss risk and fixes exactness really, with false alarm prevention, cause big Family it is flurried;The timely preservation to children loss field data is also achieved simultaneously, is gone wrong if monitored children are implicitly present in away Danger, then can quickly find lost children based on the children loss field data preserved, substantially increase children's Safety.
Specifically, determining that the children can be by cloud point with the presence or absence of the operation for risk of wandering away based on the analysis result It analyses server to execute, can also be executed by alarm system.If determining that the children whether there is based on the analysis result to wander away The operation of risk is executed by cloud Analysis server, then when determine current child exist wander away risk when, cloud Analysis server Alarm signal need to be sent to alarm system, so that alarm system and alarm.If determining the children based on the analysis result Operation with the presence or absence of risk of wandering away is executed by alarm system, and cloud Analysis server needs that the monitoring video flow will be carried out deep The analysis result that degree analysis obtains is sent to alarm system, carries out logic judgment and alarm decision by alarm system.
Step 150 is alarmed.
If the children being monitored have risk of wandering away, alarm, to improve the related monitoring such as children parent, relative The attention of people.
Children loss prevention method provided in this embodiment, by the monitoring video flow for obtaining target scene;Based on the prison Whether control video flowing determines in the target scene with the presence of children;If it is determined that with the presence of children in the target scene, then it is right The monitoring video flow carries out depth analysis, obtains analysis result;Determine that the children whether there is based on the analysis result It wanders away risk;If the children have risk of wandering away, the technological means alarmed is realized to the institute in wide area There are children to carry out the purpose of anti-monitoring of wandering away, when children, which exist, wanders away risk, by the target scene that will include the children Monitoring video flow is saved, and is sent to the corresponding terminal device of Security Personnel, to carry out manual review, is improved children and is walked It goes wrong and nearly fixes exactness really, while also achieving the timely preservation to children loss field data, help to find away as early as possible Lose children.
Embodiment two
Fig. 2 is a kind of flow diagram of children loss prevention method provided by Embodiment 2 of the present invention.In above-described embodiment On the basis of, the present embodiment " determines in the target scene whether have children to deposit based on the monitoring video flow to step 120 " be optimized, it realizes particular by combination face recognition technology and picture recognition technology to children in target scene Identification.Referring specifically to shown in Fig. 2, described method includes following steps:
Step 210, the monitoring video flow for obtaining target scene.
Step 220 extracts an at least frame original image from the monitoring video flow.
Step 230, for each frame original image, present frame original image is input to preset Face datection model, is obtained The co-ordinate position information and dimension information of each face into present frame original image.
Specifically, sequentially inputting every frame original image to preset Face datection model, Face datection model is for every Frame original image carries out operation, obtains the co-ordinate position information and dimension information for each face that every frame original image includes, In, the co-ordinate position information of the face refers specifically to the co-ordinate position information of face key point, such as the coordinate bit of two eyes The co-ordinate position information of confidence breath, the co-ordinate position information of nose and mouth, the dimension information refer specifically to facial contour Dimension information.
Wherein, the Face datection model is fabricated based on deep learning algorithm, by utilizing a large amount of labeled people Face picture carries out training in advance and obtains, the labeled face picture refer to be marked in picture face co-ordinate position information and The picture of dimension information so that Face datection model learning and remember identification face characteristic information calculating process, reaching can Identify the co-ordinate position information of face and the effect of dimension information in various pictures.When a picture is admitted to trained people When face detection model, Face datection model detects in the picture whether have face automatically, if detecting in the picture there is face, into One step detects the co-ordinate position information and dimension information of face, and the co-ordinate position information and dimension information for the face that will test Output.
Step 240, co-ordinate position information and dimension information for each face, the coordinate bit confidence according to current face The interception from corresponding frame original image of breath and dimension information includes the local picture of current face.
Wherein, the co-ordinate position information for the co-ordinate position information and dimension information of each face, according to current face And the essence of the dimension information local picture of interception comprising current face from corresponding frame original image are as follows: be based on each face Co-ordinate position information and dimension information intercept local picture respectively from corresponding frame original image, i.e., in every local picture Including a face, the corresponding frame original image refers to the picture comprising each face.
Every original image is indicated by a large amount of coordinate data point, when obtaining the co-ordinate position information of current face And when dimension information, then can from a large amount of number of coordinates strong point will belong to current face data point take out, realize from The process of local picture of the interception comprising current face in corresponding frame original image.
Each local picture is input to preset age identification model by step 250, is obtained in each local picture The corresponding age numerical value of face.
Wherein, each local picture is input to preset age identification model, obtains people in each local picture The essence of the corresponding age numerical value of face are as follows: sequentially input every local picture to preset age identification model, age Identification model carries out operation to part picture described in every respectively, obtains the age numerical value of face in every local picture.
The preset age identification model is based on deep learning algorithm and is fabricated, a large amount of labeled by utilizing Face picture is trained to obtain, and the labeled face picture, which refers to, is marked the corresponding age number of face in face picture According to picture.In order to improve the recognition accuracy of model, different types of face picture can be acquired, such as acquisition includes not the same year The face picture of age section face, acquisition includes the face picture of multiple faces, acquires the face picture etc. under different light.When new Face picture, that is, local picture when being input to age identification model, age identification model export locally after operation The corresponding age numerical value of face in picture.
Whether step 260 is determined in the target scene based on the age numerical value with the presence of children, however, it is determined that the mesh It marks in scene with the presence of children, thens follow the steps 270.
Specifically, being less than the age numerical value of preset value if it exists, it is determined that optional with the presence of children in the target scene , the preset value for example can be 10.
Step 270 carries out depth analysis to the monitoring video flow, obtains analysis result.
Step 280 determines the children based on the analysis result with the presence or absence of risk of wandering away, if the children exist It goes wrong danger, then alarms.
On the basis of above-described embodiment technical solution, the technical solution of the present embodiment, by from the monitoring video flow Present frame original image is input to preset Face datection for each frame original image by a middle extraction at least frame original image Model obtains the co-ordinate position information and dimension information of each face in present frame original image, for the coordinate bit of each face Confidence breath and dimension information, co-ordinate position information and dimension information according to current face are cut from corresponding frame original image The local picture comprising current face is taken, each local picture is input to preset age identification model, is obtained each described Whether the corresponding age numerical value of face in local picture, determined in the target scene based on the age numerical value with the presence of children Technological means, realize to the purpose whether being determined with the presence of children in target scene.
Embodiment three
Fig. 3 is a kind of flow diagram for children loss prevention method that the embodiment of the present invention three provides.In above-described embodiment On the basis of, the present embodiment continues " to determine in the target scene whether there is children based on the monitoring video flow to step 120 In the presence of " optimize, another implementation is given, is realized particular by the clothes classification of portrait in identification picture to mesh Mark the identification of children in scene.Referring specifically to shown in Fig. 3, described method includes following steps:
Step 310, the monitoring video flow for obtaining target scene.
Step 320 extracts an at least frame original image from the monitoring video flow.
Step 330, for each frame original image, present frame original image is input to preset stingy graph model, is worked as The clothes sub-pictures of each portrait Garment region in previous frame original image.
Specifically, sequentially inputting every frame original image to preset stingy graph model, scratches graph model and be directed to every frame original graph Piece carries out operation, obtains the clothes word picture for each portrait Garment region that every frame original image includes, wherein every clothes subgraph It include the Garment region of a portrait in piece.
Wherein, the stingy graph model is fabricated based on deep learning algorithm, by utilizing a large amount of labeled portrait figures Piece carries out training in advance and obtains, and the labeled portrait picture refers to the picture for being marked portrait Garment region in picture, with Make to scratch the calculating process that graph model learns and remembers portrait Garment region in identification picture, reaching can identify in various pictures The effect of portrait Garment region.When a picture is admitted to trained stingy graph model, scratches graph model and detect the picture automatically In portrait Garment region, and the portrait Garment region that will test exports to obtain the clothes subgraph including portrait Garment region Piece.
Step 340, for each clothes sub-pictures, current garment sub-pictures are input to preset dress ornament classification and are identified Model obtains the classification of clothes in current garment sub-pictures.
Wherein, for each clothes sub-pictures, current garment sub-pictures is input to preset dress ornament classification and identify mould Type obtains the essence of the classification of clothes in current garment sub-pictures are as follows: sequentially inputs every clothes sub-pictures to preset clothes Classification identification model is adornd, dress ornament classification identification model carries out operation to clothes sub-pictures described in every respectively, obtains every clothes The classification of clothes in sub-pictures, the classification of the clothes include children's garment and adult dress.
The preset dress ornament classification identification model is based on deep learning algorithm and is fabricated, and is largely marked by utilizing The clothes picture crossed is trained to obtain, and the labeled clothes picture refers to the figure for being marked clothes classification in clothes picture Piece.In order to improve the recognition accuracy of model, different types of clothes picture can be acquired, such as acquisition includes different age group The clothes picture that people is worn, acquisition include the clothes picture of different-style clothes, acquire the clothes picture etc. under different light.When When new clothes picture, that is, clothes sub-pictures are input to dress ornament classification identification model, dress ornament classification identification model is by fortune The classification that clothes in clothes sub-pictures is exported after calculation is adult dress or children's dress.
Whether step 350 is determined in the target scene based on the classification of the clothes with the presence of children, however, it is determined that described With the presence of children in target scene, 360 are thened follow the steps.
Specifically, if dress ornament classification identification model identifies that the clothes classification in each clothes sub-pictures is adult clothing Clothes, it is determined that do not have children's presence in target scene, if dress ornament classification identification model identifies in each clothes sub-pictures There are children clothes in clothes classification, it is determined that with the presence of children in target scene.
Step 360 carries out depth analysis to the monitoring video flow, obtains analysis result.
Step 370 determines the children based on the analysis result with the presence or absence of risk of wandering away, if the children exist It goes wrong danger, then alarms.
On the basis of above-described embodiment technical solution, the technical solution of the present embodiment, by from the monitoring video flow A middle extraction at least frame original image;For each frame original image, present frame original image is input to preset stingy graph model, Obtain the clothes sub-pictures of each portrait Garment region in present frame original image;For each clothes sub-pictures, by current clothing It takes sub-pictures and is input to preset dress ornament classification identification model, obtain the classification of clothes in current garment sub-pictures;Based on described Whether the classification of clothes determines technological means existing for children in the target scene, whether realize has in target scene There is the purpose being determined in children.
Example IV
Fig. 4 is a kind of flow diagram for children loss prevention method that the embodiment of the present invention four provides.In above-described embodiment On the basis of, the present embodiment continues " to determine in the target scene whether there is children based on the monitoring video flow to step 120 In the presence of " optimize, another implementation is given, specifically by combining face recognition technology and picture recognition technology It determines on the basis of having existing for children in target scene, by dress ornament classification recognition and verification, whether the above-mentioned children determined are true Be children, the benefit optimized in this way be improve children identification accuracy.Referring specifically to shown in Fig. 4, the method includes Following steps:
Step 410, the monitoring video flow for obtaining target scene.
Step 420 extracts an at least frame original image from the monitoring video flow.
Step 430, for each frame original image, present frame original image is input to preset Face datection model, is obtained The co-ordinate position information and dimension information of each face into present frame original image.
Step 440, co-ordinate position information and dimension information for each face, the coordinate bit confidence according to current face The interception from corresponding frame original image of breath and dimension information includes the local picture of current face.
Each local picture is input to preset age identification model by step 450, is obtained in each local picture The corresponding age numerical value of face.
Whether step 460 is determined in the target scene based on the age numerical value with the presence of children, however, it is determined that the mesh In the presence of there are children in mark scene, step 470 is continued to execute.
Step 470, based on the corresponding age numerical value of face in each local picture in original image respectively to children Corresponding portrait and the corresponding portrait of adult are marked.
Specifically, for example, then will if the corresponding age numerical value of face is greater than preset value, such as 15 in current part picture Currently picture corresponding portrait in part is labeled as adult in original image.If the corresponding age number of face in current part picture Value is less than preset value, then picture corresponding portrait in part current in original image is labeled as children.
Step 480, the clothes sub-pictures for obtaining the children's portrait Garment region being labeled in original image.
The clothes sub-pictures are input to preset dress ornament classification identification model by step 490, obtain the clothes subgraph The classification of clothes in piece.
Whether the labeled children's portrait of step 4100, the classification confirmation based on the clothes is genuine children's portrait, if Labeled children's portrait is genuine children's portrait, then continues to execute step 4110.
If being marked as the clothes classification of children's portrait as children's dress, it is determined that labeled children's portrait is genuine youngster Virgin portrait, if being marked as the clothes classification of children's portrait as adult dress, it is determined that labeled children's portrait is not genuine Children's portrait.
Step 4110 carries out depth analysis to the monitoring video flow, obtains analysis result.
Step 4120 determines the children based on the analysis result with the presence or absence of risk of wandering away, if the children exist It wanders away risk, then alarms.
On the basis of above-described embodiment technical solution, the technical solution of the present embodiment, by combining recognition of face skill Art and picture recognition technology are determined on the basis of having existing for children in target scene, by the confirmation of dress ornament classification identification technology State the children that determine whether be really children technological means, improve the accuracy identified to children in target scene.
Embodiment five
Fig. 5 is a kind of flow diagram for children loss prevention method that the embodiment of the present invention five provides.In above-described embodiment On the basis of, the present embodiment is to step 130 " if it is determined that with the presence of children in the target scene, then to the monitoring video flow Depth analysis is carried out, analysis result is obtained " it is optimized, monitoring is regarded particular by being realized in conjunction with depth learning technology The identification of face emotional state in frequency stream.Referring specifically to shown in Fig. 5, described method includes following steps:
Step 510, the monitoring video flow for obtaining target scene.
Whether step 520 is determined in the target scene based on the monitoring video flow with the presence of children, however, it is determined that described With the presence of children in target scene, 530 are thened follow the steps.
Step 530 extracts the continuous original image of at least two frames from the monitoring video flow.
Step 540, for each frame original image, present frame original graph is identified by recognition of face key point extractive technique Each children's face presets the profile coordinate information of organ in piece.
Specifically, identifying the face in each frame original image first with face recognition technology, it is then based on picture recognition Technology identifies children's face in face, identifies that each children's face presets organ finally by recognition of face key point extractive technique Profile coordinate information, the default organ includes eyes, mouth and nose etc..
Step 550, according to sequential relationship, children's face each in each frame original image is preset to the profile coordinate information of organ It connects.
Specifically, according to the sequencing of time, by the default organ of children's face same in each frame original image Profile coordinate information is connected, then by the information input after the series connection of each children of correspondence to preset Emotion identification mould Type obtains the corresponding emotional state of each children's face, wherein the emotional state includes crying and screaming and anxiety.By to continuous Children's face presets the analysis of the profile coordinate information variation of organ in multiframe original image in a period of time, and it is accurate to can get Spend higher emotional state analysis result.
Step 560, by the information input after series connection to preset Emotion identification model, it is corresponding to obtain each children's face Emotional state.
Specifically, the information input after the corresponding series connection of each children's face is obtained to preset Emotion identification model The corresponding emotional state of each children's face.The preset Emotion identification model is based on deep learning algorithm and is fabricated, and passes through It is trained to obtain using a large amount of labeled face picture sequence frame, the labeled face picture sequence frame refers to It is marked the continuous multiframe picture of the corresponding emotional state of face.In order to improve the recognition accuracy of model, difference can be acquired The face picture sequence frame of emotional state, such as acquisition emotional state are the face picture sequence frame of anxious children, acquire feelings Not-ready status is face picture sequence frame of the children to cry and scream etc..When new face picture sequence frame is input to Emotion identification model When, Emotion identification model exports the corresponding emotional state of face in face picture sequence frame after operation.
Step 570, when the emotional state of current child be cry and scream or anxiety, and within the scope of current child pre-determined distance When there is no adult, determines that current child has risk of wandering away, alarm.
On the basis of above-described embodiment technical solution, the technical solution of the present embodiment, by former based on continuous multiframe Beginning picture carries out the identification of face emotional state, improves the accuracy of identification of face emotional state, aids in determining whether monitored youngster Child realizes the anti-monitoring of wandering away to children in large area target scene with the presence or absence of risk of wandering away.
Embodiment six
Fig. 6 is a kind of flow diagram for children loss prevention method that the embodiment of the present invention six provides.In above-described embodiment On the basis of, the present embodiment is to step 130 " if it is determined that with the presence of children in the target scene, then to the monitoring video flow Depth analysis is carried out, analysis result is obtained " continue optimization, it identifies in picture and grows up particular by picture recognition technology Similarity between face and children's face is sentenced with determining that the adult and the children whether there is genetic connection for assisting The fixed children improve the judgement accuracy of children loss risk with the presence or absence of risk of wandering away.It is described shown in referring specifically to fig. 6 Method includes the following steps:
Step 610, the monitoring video flow for obtaining target scene.
Whether step 620 is determined in the target scene based on the monitoring video flow with the presence of children, however, it is determined that described With the presence of children in target scene, 630 are thened follow the steps.
Step 630 extracts the continuous original image of at least two frames from the monitoring video flow.
Step 640, for each frame original image, present frame original graph is identified by recognition of face key point extractive technique Each children's face presets the profile coordinate information of organ in piece.
Step 650, according to sequential relationship, children's face each in each frame original image is preset to the profile coordinate information of organ It connects.
Step 660, by the information input after series connection to preset Emotion identification model, it is corresponding to obtain each children's face Emotional state.
Wherein, the emotional state includes crying and screaming and anxiety.
Step 670, for each frame original image, determine in present frame original image and grow up between face and children's face Similarity.
Specifically, similarity calculation can be constructed by depth learning technology, pass through similarity calculation realization pair The calculating for the similarity between face and children's face of growing up in original image.Similarity calculation can be by being largely labeled Picture be trained, which refers to the picture for being marked in picture similarity between face, by similar Degree computation model is trained, so that model learning and remembering the process for calculating similarity between face in picture.When new figure When piece is input to trained similarity calculation, model carries out similarity calculation automatically and exports similarity result.
Step 680 determines between adult and children according to the similarity with the presence or absence of genetic connection.
Specifically, if the similarity is greater than given threshold, it is determined that there are genetic connection, institutes between adult and children Stating threshold value for example can be 80%.
Step 690, when the emotional state of current child be cry and scream or anxiety, and within the scope of current child pre-determined distance When having adult, but genetic connection being not present between the adult and current child, determine that current child has risk of wandering away, It alarms.
On the basis of above-described embodiment technical solution, the technical solution of the present embodiment, by further calculate children with Human face similarity degree around it between adult, determine children around with the presence or absence of with the children have genetic connection relatives, And then the auxiliary judgement children improve the judgement precision for risk of wandering away, realize to target scene with the presence or absence of risk of wandering away The security monitoring of middle children.
Embodiment seven
Fig. 7 is a kind of structural schematic diagram for children loss prevention device that the embodiment of the present invention seven provides.It is shown in Figure 7, Described device includes: to obtain module 710, children's determining module 720, analysis module 730, risk determining module 740 and alarm mould Block 750;
Wherein, module 710 is obtained, for obtaining the monitoring video flow of target scene;Children's determining module 720 is used for base It whether determines in the target scene in the monitoring video flow with the presence of children;Analysis module 730, for if it is determined that the mesh It marks in scene with the presence of children, then depth analysis is carried out to the monitoring video flow, obtain analysis result;Risk determining module 740, for determining the children with the presence or absence of risk of wandering away based on the analysis result;Alarm module 750, if being used for the youngster There is risk of wandering away in child, then alarm.
Further, children's determining module 720 includes: extraction unit, for extracting at least from the monitoring video flow One frame original image;Face identification unit, for for each frame original image, present frame original image to be input to preset people Face detection model obtains the co-ordinate position information and dimension information of each face in present frame original image;Interception unit is used for For the co-ordinate position information and dimension information of each face, according to current face co-ordinate position information and dimension information from Interception includes the local picture of current face in corresponding frame original image;Age recognition unit is used for each local picture It is input to preset age identification model, obtains the corresponding age numerical value of face in each local picture;Determination unit is used for It whether is determined in the target scene based on the age numerical value with the presence of children.
Further, described device further include: mark module determines in the target scene for working as with the presence of children When, based on the corresponding age numerical value of face in each local picture in original image respectively to the corresponding portrait of children and The corresponding portrait of adult is marked.
Further, children's determining module 720 further include: clothes sub-pictures acquiring unit, for when the determining target In the presence of there are children in scene, the clothes sub-pictures for the children's portrait Garment region being labeled in original image are obtained;Clothes class Other recognition unit obtains the clothes subgraph for the clothes sub-pictures to be input to preset dress ornament classification identification model The classification of clothes in piece;Confirmation unit, whether the children portrait labeled for the classification confirmation based on the clothes is genuine Children's portrait;Wherein, the classification of the clothes includes children's garment and adult dress.
Further, children's determining module 720 further include: scratch module, be used for for each frame original image, by present frame Original image is input to preset stingy graph model, obtains the clothes sub-pictures of each portrait Garment region in present frame original image; Determination unit, for whether being determined in the target scene based on the classification of the clothes with the presence of children.
Further, analysis module 730 includes: picture extraction unit, for extracting at least from the monitoring video flow The continuous original image of two frames;Face key point recognition unit, for passing through recognition of face key point for each frame original image Extractive technique identifies the profile coordinate information of the default organ of each children's face in present frame original image;Series unit is used for According to sequential relationship, the profile coordinate information that children's face each in each frame original image presets organ is connected;Input is single Member obtains the corresponding emotional state of each children's face for the information input after connecting to preset Emotion identification model; Wherein, the emotional state includes crying and screaming and anxiety.
Further, risk determining module 740 is specifically used for: when the emotional state of current child is to cry and scream or anxiety, And when with there is no adult within the scope of current child pre-determined distance, determine that current child has risk of wandering away.
Further, analysis module 730 further include: similarity determining unit, for for each frame original image, determination to be worked as The similarity grown up between face and children's face in previous frame original image;Genetic connection determination unit, for according to the phase It determines between adult and children like spending with the presence or absence of genetic connection.Corresponding, risk determining module 740 is specifically used for: when working as The emotional state of preceding children be cry and scream or anxiety, and with have adult within the scope of current child pre-determined distance, but the adult When genetic connection being not present between people and current child, determine that current child has risk of wandering away.
Further, described device further includes preserving module, if there is risk of wandering away for the children, by the mesh The monitoring video flow of mark scene is saved, and is sent to the corresponding terminal device of Security Personnel, to carry out manual review.
Children loss prevention device provided in this embodiment, by the monitoring video flow for obtaining target scene;Based on the prison Whether control video flowing determines in the target scene with the presence of children;If it is determined that with the presence of children in the target scene, then it is right The monitoring video flow carries out depth analysis, obtains analysis result;Determine that the children whether there is based on the analysis result It wanders away risk;If the children have risk of wandering away, the technological means alarmed is realized to the institute in wide area There are children to carry out the purpose of anti-monitoring of wandering away, when children, which exist, wanders away risk, by the target scene that will include the children Monitoring video flow is saved, and is sent to the corresponding terminal device of Security Personnel, to carry out manual review, is improved children and is walked It goes wrong and nearly fixes exactness really, while also achieving the timely preservation to children loss field data, help to find away as early as possible Lose children.
Anti- youngster provided by any embodiment of the invention can be performed in children loss prevention device provided by the embodiment of the present invention Child wanders away method, has the corresponding functional module of execution method and beneficial effect, not the skill of detailed description in the above-described embodiments Art details, reference can be made to children loss prevention method provided by any embodiment of the invention.
Embodiment eight
Fig. 8 is the structural schematic diagram for a kind of electronic equipment that the embodiment of the present invention eight provides.Fig. 8, which is shown, to be suitable for being used in fact The block diagram of the example devices 12 of existing embodiment of the present invention.The equipment 12 that Fig. 8 is shown is only an example, should not be to this hair The function and use scope of bright embodiment bring any restrictions.
As shown in figure 8, equipment 12 is showed in the form of universal computing device.The component of equipment 12 may include but unlimited In one or more processor or processing unit 16, system storage 28, connecting different system components, (including system is deposited Reservoir 28 and processing unit 16) bus 18.
Bus 18 indicates one of a few class bus structures or a variety of, including memory bus or Memory Controller, Peripheral bus, graphics acceleration port, processor or the local bus using any bus structures in a variety of bus structures.It lifts For example, these architectures include but is not limited to industry standard architecture (ISA) bus, microchannel architecture (MAC) Bus, enhanced isa bus, Video Electronics Standards Association (VESA) local bus and peripheral component interconnection (PCI) bus.
Equipment 12 typically comprises a variety of computer system readable media.These media can be it is any can be by equipment 12 The usable medium of access, including volatile and non-volatile media, moveable and immovable medium.
System storage 28 may include the computer system readable media of form of volatile memory, such as arbitrary access Memory (RAM) 30 and/or cache memory 32.Equipment 12 may further include it is other it is removable/nonremovable, Volatile/non-volatile computer system storage medium.Only as an example, storage system 34 can be used for reading and writing irremovable , non-volatile magnetic media (Fig. 8 do not show, commonly referred to as " hard disk drive ").Although being not shown in Fig. 8, use can be provided In the disc driver read and write to removable non-volatile magnetic disk (such as " floppy disk "), and to removable anonvolatile optical disk The CD drive of (such as CD-ROM, DVD-ROM or other optical mediums) read-write.In these cases, each driver can To be connected by one or more data media interfaces with bus 18.Memory 28 may include at least one program product, The program product has one group of (such as acquisition module 710, children's determining module 720, analysis module of children loss prevention device 730, risk determining module 740 and alarm module 750) program module, these program modules are configured to perform each reality of the present invention Apply the function of example.
With one group (such as the acquisition module 710 of children loss prevention device, children's determining module 720, analysis module 730, Risk determining module 740 and alarm module 750) program module 42 program/utility 40, can store in such as memory In 28, such program module 42 include but is not limited to operating system, one or more application program, other program modules with And program data, it may include the realization of network environment in each of these examples or certain combination.Program module 42 is logical Often execute the function and/or method in embodiment described in the invention.
Equipment 12 can also be communicated with one or more external equipments 14 (such as keyboard, sensing equipment, display 24 etc.), Can also be enabled a user to one or more equipment interacted with the equipment 12 communication, and/or with enable the equipment 12 with One or more of the other any equipment (such as network interface card, modem etc.) communication for calculating equipment and being communicated.It is this logical Letter can be carried out by input/output (I/O) interface 22.Also, equipment 12 can also by network adapter 20 and one or The multiple networks of person (such as local area network (LAN), wide area network (WAN) and/or public network, such as internet) communication.As shown, Network adapter 20 is communicated by bus 18 with other modules of equipment 12.It should be understood that although not shown in the drawings, can combine Equipment 12 use other hardware and/or software module, including but not limited to: microcode, device driver, redundant processing unit, External disk drive array, RAID system, tape drive and data backup storage system etc..
Processing unit 16 by the program that is stored in system storage 28 of operation, thereby executing various function application and Data processing, such as realize children loss prevention method provided by the embodiment of the present invention.
Embodiment nine
The embodiment of the present invention nine additionally provides a kind of computer readable storage medium, is stored thereon with computer program, should Children loss prevention method described in above-described embodiment is realized when program is executed by processor, this method comprises:
Obtain the monitoring video flow of target scene;
It whether is determined in the target scene based on the monitoring video flow with the presence of children;
If it is determined that then carrying out depth analysis with the presence of children in the target scene to the monitoring video flow, being divided Analyse result;
Determine the children with the presence or absence of risk of wandering away based on the analysis result;
If the children have risk of wandering away, alarm.
The computer storage medium of the embodiment of the present invention, can be using any of one or more computer-readable media Combination.Computer-readable medium can be computer-readable signal media or computer readable storage medium.It is computer-readable Storage medium for example may be-but not limited to-the system of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, device or Device, or any above combination.The more specific example (non exhaustive list) of computer readable storage medium includes: tool There are electrical connection, the portable computer diskette, hard disk, random access memory (RAM), read-only memory of one or more conducting wires (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD- ROM), light storage device, magnetic memory device or above-mentioned any appropriate combination.In this document, computer-readable storage Medium can be any tangible medium for including or store program, which can be commanded execution system, device or device Using or it is in connection.
Computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal, Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for By the use of instruction execution system, device or device or program in connection.
The program code for including on computer-readable medium can transmit with any suitable medium, including --- but it is unlimited In wireless, electric wire, optical cable, RF etc. or above-mentioned any appropriate combination.
The computer for executing operation of the present invention can be write with one or more programming languages or combinations thereof Program code, described program design language include object oriented program language-such as Java, Smalltalk, C++, Further include conventional procedural programming language-such as " C " language or similar programming language.Program code can be with It fully executes, partly execute on the user computer on the user computer, being executed as an independent software package, portion Divide and partially executes or executed on a remote computer or server completely on the remote computer on the user computer.? Be related in the situation of remote computer, remote computer can pass through the network of any kind --- including local area network (LAN) or Wide area network (WAN)-be connected to subscriber computer, or, it may be connected to outer computer (such as mentioned using Internet service It is connected for quotient by internet).
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation, It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.

Claims (13)

1. a kind of children loss prevention method characterized by comprising
Obtain the monitoring video flow of target scene;
It whether is determined in the target scene based on the monitoring video flow with the presence of children;
If it is determined that then carrying out depth analysis to the monitoring video flow with the presence of children in the target scene, analysis knot is obtained Fruit;
Determine the children with the presence or absence of risk of wandering away based on the analysis result;
If the children have risk of wandering away, alarm.
2. the method according to claim 1, wherein being determined in the target scene based on the monitoring video flow Whether with the presence of children, comprising:
An at least frame original image is extracted from the monitoring video flow;
For each frame original image, present frame original image is input to preset Face datection model, it is original to obtain present frame The co-ordinate position information and dimension information of each face in picture;
Co-ordinate position information and size letter for the co-ordinate position information and dimension information of each face, according to current face Cease the local picture that the interception from corresponding frame original image includes current face;
Each local picture is input to preset age identification model, obtains face corresponding year in each local picture Age numerical value;
It whether is determined in the target scene based on the age numerical value with the presence of children.
3. according to the method described in claim 2, it is characterized in that, being gone back in the presence of having children in the determining target scene Include:
Based on the corresponding age numerical value of face in each local picture in original image respectively to the corresponding portrait of children with And the corresponding portrait of adult is marked.
4. according to the method described in claim 3, it is characterized in that, being gone back in the presence of having children in the determining target scene Include:
Obtain the clothes sub-pictures for the children's portrait Garment region being labeled in original image;
The clothes sub-pictures are input to preset dress ornament classification identification model, obtain the class of clothes in the clothes sub-pictures Not;
Whether the labeled children's portrait of the classification confirmation based on the clothes is genuine children's portrait;
Wherein, the classification of the clothes includes children's garment and adult dress.
5. the method according to claim 1, wherein being determined in the target scene based on the monitoring video flow Whether with the presence of children, comprising:
An at least frame original image is extracted from the monitoring video flow;
For each frame original image, present frame original image is input to preset stingy graph model, obtains present frame original image In each portrait Garment region clothes sub-pictures;
For each clothes sub-pictures, current garment sub-pictures are input to preset dress ornament classification identification model, are worked as The classification of clothes in preceding clothes sub-pictures;
It whether is determined in the target scene based on the classification of the clothes with the presence of children.
6. method according to any one of claims 1-5, which is characterized in that carry out depth point to the monitoring video flow Analysis obtains analysis result, comprising:
For each frame original image in the monitoring video flow in the continuous original image of at least two frames, closed by recognition of face Key point extractive technique identifies the profile coordinate information of the default organ of each children's face in present frame original image;
According to sequential relationship, the profile coordinate information that children's face each in each frame original image presets organ is connected;
By the information input after series connection to preset Emotion identification model, the corresponding emotional state of each children's face is obtained;
Wherein, the emotional state includes crying and screaming and anxiety.
7. according to the method described in claim 6, it is characterized in that, determining that the children whether there is based on the analysis result It wanders away risk, comprising:
When the emotional state of current child be cry and scream or anxiety, and within the scope of current child pre-determined distance without adult When, determine that current child has risk of wandering away.
8. according to the method described in claim 6, it is characterized in that, being divided monitoring video flow progress depth analysis Analyse result, further includes:
For each frame original image, the similarity grown up between face and children's face in present frame original image is determined;
It is determined between adult and children according to the similarity with the presence or absence of genetic connection.
9. according to the method described in claim 8, it is characterized in that, determining that the children whether there is based on the analysis result It wanders away risk, comprising:
When the emotional state of current child be cry and scream or anxiety, and with have adult within the scope of current child pre-determined distance, but It is that there is no when genetic connection, determine that current child has risk of wandering away between the adult and current child.
10. method according to claim 1-5, which is characterized in that further include: if the children exist to walk to go wrong Danger, then save the monitoring video flow of the target scene, and be sent to the corresponding terminal device of Security Personnel, to carry out Manual review.
11. a kind of children loss prevention device characterized by comprising
Module is obtained, for obtaining the monitoring video flow of target scene;
Children's determining module, for whether being determined in the target scene based on the monitoring video flow with the presence of children;
Analysis module, for if it is determined that then carrying out depth point to the monitoring video flow with the presence of children in the target scene Analysis obtains analysis result;
Risk determining module, for determining the children with the presence or absence of risk of wandering away based on the analysis result;
Alarm module is alarmed if there is risk of wandering away for the children.
12. a kind of electronic equipment, which is characterized in that the electronic equipment includes:
One or more processors;
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real Now such as children loss prevention method of any of claims 1-10.
13. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor Such as children loss prevention method of any of claims 1-10 is realized when execution.
CN201910172200.8A 2019-03-07 2019-03-07 A kind of children loss prevention method, apparatus, electronic equipment and storage medium Pending CN109887234A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910172200.8A CN109887234A (en) 2019-03-07 2019-03-07 A kind of children loss prevention method, apparatus, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910172200.8A CN109887234A (en) 2019-03-07 2019-03-07 A kind of children loss prevention method, apparatus, electronic equipment and storage medium

Publications (1)

Publication Number Publication Date
CN109887234A true CN109887234A (en) 2019-06-14

Family

ID=66931199

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910172200.8A Pending CN109887234A (en) 2019-03-07 2019-03-07 A kind of children loss prevention method, apparatus, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN109887234A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110446015A (en) * 2019-08-30 2019-11-12 北京青岳科技有限公司 A kind of abnormal behaviour monitoring method based on computer vision and system
CN110718041A (en) * 2019-09-18 2020-01-21 恒大智慧科技有限公司 Method, device and system for preventing children from getting lost and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004151820A (en) * 2002-10-29 2004-05-27 Hitachi Eng Co Ltd Lost child searching and monitoring system
CN107169461A (en) * 2017-05-19 2017-09-15 苏州寅初信息科技有限公司 A kind of intelligent and safe method for tracing and its system for market
CN107392182A (en) * 2017-08-17 2017-11-24 宁波甬慧智能科技有限公司 A kind of face collection and recognition method and device based on deep learning
CN108460324A (en) * 2018-01-04 2018-08-28 上海孩子通信息科技有限公司 A method of child's mood for identification
CN109035686A (en) * 2018-07-10 2018-12-18 北京三快在线科技有限公司 A kind of pre- anti-lost alarm method and device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004151820A (en) * 2002-10-29 2004-05-27 Hitachi Eng Co Ltd Lost child searching and monitoring system
CN107169461A (en) * 2017-05-19 2017-09-15 苏州寅初信息科技有限公司 A kind of intelligent and safe method for tracing and its system for market
CN107392182A (en) * 2017-08-17 2017-11-24 宁波甬慧智能科技有限公司 A kind of face collection and recognition method and device based on deep learning
CN108460324A (en) * 2018-01-04 2018-08-28 上海孩子通信息科技有限公司 A method of child's mood for identification
CN109035686A (en) * 2018-07-10 2018-12-18 北京三快在线科技有限公司 A kind of pre- anti-lost alarm method and device

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
瞿道庆 等: "结合深度度量学习的血缘关系识别", 《中国图象图形学报》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110446015A (en) * 2019-08-30 2019-11-12 北京青岳科技有限公司 A kind of abnormal behaviour monitoring method based on computer vision and system
CN110718041A (en) * 2019-09-18 2020-01-21 恒大智慧科技有限公司 Method, device and system for preventing children from getting lost and storage medium

Similar Documents

Publication Publication Date Title
US10402632B2 (en) Pose-aligned networks for deep attribute modeling
Bolanos et al. Toward storytelling from visual lifelogging: An overview
Ma et al. Depth-based human fall detection via shape features and improved extreme learning machine
Jalal et al. A Depth Video-based Human Detection and Activity Recognition using Multi-features and Embedded Hidden Markov Models for Health Care Monitoring Systems.
US10679047B2 (en) System and method for pose-aware feature learning
Bian et al. Fall detection based on body part tracking using a depth camera
US9792489B2 (en) Time-in store estimation using facial recognition
JP6022681B2 (en) Interactive clothing search in online store
US20160078289A1 (en) Gesture Recognition Apparatuses, Methods and Systems for Human-Machine Interaction
TWI505200B (en) Method,system,computer program product and process for determining parts and associated attributes of an object in video
Dehghan et al. Dager: Deep age, gender and emotion recognition using convolutional neural network
JP6045549B2 (en) Method and system for recognizing emotions and actions
JP4509209B2 (en) Hierarchical system for object recognition in images
Dhiman et al. A review of state-of-the-art techniques for abnormal human activity recognition
JP6850291B2 (en) Methods and devices for minimizing false positives in face recognition applications
Cardinaux et al. Video based technology for ambient assisted living: A review of the literature
US10271017B2 (en) System and method for generating an activity summary of a person
Brulin et al. Posture recognition based on fuzzy logic for home monitoring of the elderly
US8295545B2 (en) System and method for model based people counting
US6128397A (en) Method for finding all frontal faces in arbitrarily complex visual scenes
Burçin et al. Down syndrome recognition using local binary patterns and statistical evaluation of the system
US9977951B2 (en) Picture ranking method, and terminal
US8503739B2 (en) System and method for using contextual features to improve face recognition in digital images
US20180314881A1 (en) Classifying facial expressions using eye-tracking cameras
CN105809144B (en) A kind of gesture recognition system and method using movement cutting

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