CN110309768A - The staff's detection method and equipment of car test station - Google Patents
The staff's detection method and equipment of car test station Download PDFInfo
- Publication number
- CN110309768A CN110309768A CN201910576535.6A CN201910576535A CN110309768A CN 110309768 A CN110309768 A CN 110309768A CN 201910576535 A CN201910576535 A CN 201910576535A CN 110309768 A CN110309768 A CN 110309768A
- Authority
- CN
- China
- Prior art keywords
- image
- personnel
- test station
- car test
- clothes
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- 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
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Abstract
The object of the present invention is to provide a kind of staff's detection method of car test station and equipment, the present invention obtains the first comparison result by the way that the facial image in the image of the car test station to be compared with staff's face template image；Clothes area image in the image of the car test station is compared with staff's clothes template image, obtains the second comparison result；Mutual alignment relation between human body key point position in the image of the car test station is compared with the mutual alignment relation between template human body key point position, obtains third comparison result；Judge that the staff of car test station detects whether qualification based on the first, second, and third comparison result, improve the accuracy rate of personnel's detection, and realize staff's identification, dressing identification and compulsory exercise identification, the whole-process automatic detection of review process simultaneously, both manpower has been saved, has in turn ensured the disclosure, just of detection work.
Description
Technical field
The present invention relates to computer field more particularly to the staff's detection methods and equipment of a kind of car test station.
Background technique
With the continuous social and economic development with the continuous improvement of living standards of the people, Urban vehicles poputation rapidly increases
It is long.The workload of automotive vehicle annual test also increases rapidly therewith.Staff's inspection of car test station in traditional vehicle annual test
It surveys mainly by artificial detection, this method cost of labor is higher, and efficiency is lower, and the operation of repeatability detection for a long time is easy to make
It obtains testing staff and generates the defective modes such as tired, absent minded, influence Detection accuracy.
How dynamic detection accurately and rapidly to be carried out to the staff of car test station, while avoiding artificial detection cost
Height, the drawbacks such as testing staff's fatiguability, easy missing error are technical problems urgently to be solved.
Summary of the invention
It is an object of the present invention to provide a kind of staff's detection method of car test station and equipment.
According to an aspect of the invention, there is provided a kind of staff's detection method of car test station, this method comprises:
Obtain the image of car test station；
Facial image in the image of the car test station is compared with staff's face template image, obtains
One comparison result；
Clothes area image in the image of the car test station is compared with staff's clothes template image, is obtained
To the second comparison result；
Mutual alignment relation between human body key point position in the image of the car test station is closed with template human body
Mutual alignment relation between key point position is compared, and obtains third comparison result；
Judge that the staff of car test station detects whether qualification based on the first, second, and third comparison result.
Further, in the above method, by the facial image and staff's face mould in the image of the car test station
Domain picture is compared, and obtains the first comparison result, comprising:
Using personnel's detection model based on deep learning network, judge in the image of the car test station with the presence or absence of people
Member's area image, personnel area image, then carry out the corresponding personnel area flag bit of the image of the car test station if it does not exist
Label；Personnel area image if it exists then uses personnel's disaggregated model based on deep learning network, judges the personnel area
It whether there is personnel in image, if it does not exist personnel, then the image of the car test station corresponding personnel identifier position be marked；
Personnel if it exists obtain the personnel area image, using the Face datection model based on deep learning network, inspection
It surveys and whether there is human face region image in the personnel area image, if it does not exist human face region image, then to the car test station
The corresponding face flag bit of image be marked；Face if it exists then obtains the human face region image, by the face area
Area image is compared with staff's face template image, obtains the first comparison result.
Further, in the above method, using personnel's detection model based on deep learning network, judge the car test work
With the presence or absence of before the image of personnel area in the image of position, further includes:
Obtain the image of the sample car test station under different illumination, different shooting angles；
Personnel position in the image of the sample car test station is marked using rectangle frame, and is labeled as personnel, is obtained
Sample image after label；
Deep neural network model is detected using the sample image training objective after the label, is based on depth to obtain
Practise personnel's detection model of network.
Further, in the above method, using personnel's disaggregated model based on deep learning network, judge the personnel area
With the presence or absence of before personnel in area image, further includes:
Using personnel's detection model based on deep learning network, obtain in the image of the sample car test station not
With the sample personnel area image of position；
Classify to sample personnel area image, is divided into two class of someone's image and unmanned image；
Using someone's image and unmanned image training objective depth of assortment neural network model, obtains and be based on depth
Practise personnel's disaggregated model of network.
Further, in the above method, using the Face datection model based on deep learning network, the personnel area is detected
With the presence or absence of before human face region image in area image, further includes:
Obtain different sample personnel area images；
Human face region image position in the image of the sample personnel area is marked using rectangle frame, and marking is people
Face；
Deep neural network model is detected using the sample personnel area image training objective after label, to obtain based on deep
Spend the Face datection model of learning network.
Further, in the above method, by the clothes area image and staff's clothes in the image of the car test station
Dress template image is compared, and obtains the second comparison result, comprising:
Using the clothes detection model based on deep learning network, judge in the image of the car test station with the presence or absence of clothes
Area image is filled, clothes area image, then be marked the corresponding clothing logos position of the image of the car test station if it does not exist；
Clothes area image if it exists then obtains the clothes area image, by the clothes area image and work people
Member's clothes template image is compared, and obtains the second comparison result.
Further, in the above method, using the clothes detection model based on deep learning network, judge the car test work
With the presence or absence of before clothes area image in the image of position, further includes:
Obtain different sample personnel area images；
Using clothes area image position in rectangle frame label sample personnel area figure, and it is labeled as clothes；
Depth neural network model is detected using the personnel area image training objective after label, is based on depth to obtain
Practise the clothes detection model of network.
It further, will be mutual between the human body key point position in the image of the car test station in the above method
Mutual alignment relation between positional relationship and template human body key point position is compared, and obtains third comparison result, comprising:
Using the human body critical point detection network model based on deep learning, human body is obtained in the personnel area image
Key point position, if human body key point position has not been obtained, key point mark corresponding to the image of the car test station
Position is marked；
If getting human body key point position, the mutual alignment relation for the human body key point position that will acquire with
Mutual alignment relation between template human body key point position is compared, and obtains third comparison result.
Further, in the above method, using the human body critical point detection network model based on deep learning, in the people
In member's area image before acquisition human body key point position, further includes:
Obtain different sample personnel area images；
Label includes the crown, neck, left shoulder, right shoulder, left hand elbow, right hand elbow, a left side in the image of each sample personnel area
The key points coordinate such as wrist, right finesse, thoracic cavity, pelvis；
Using the crucial Point Target Detection deep neural network model of sample personnel area image training after label, to obtain
Human body critical point detection network model based on deep learning.
Further, in the above method, the staff of car test station is judged based on the first, second, and third comparison result
Detect whether qualification, comprising:
Based on the personnel area flag bit, personnel identifier position, face flag bit, clothing logos position and described first,
Two and third comparison result, judge that the staff of car test station detects whether qualification.
According to another aspect of the present invention, a kind of staff's detection device of car test station is additionally provided, the equipment packet
It includes:
Acquisition device, for obtaining the image of car test station；
First comparison device, for by the image of the car test station facial image and staff's face template figure
As being compared, the first comparison result is obtained；
Second comparison device, for by the image of the car test station clothes area image and staff's clothes mould
Domain picture is compared, and obtains the second comparison result；
Third comparison device, for by the mutual alignment between the human body key point position in the image of the car test station
Mutual alignment relation between relationship and template human body key point position is compared, and obtains third comparison result；
Judgment means, for judged based on the first, second, and third comparison result car test station staff detect be
No qualification.
According to another aspect of the present invention, a kind of equipment based on calculating is also provided characterized by comprising
Processor；And
It is arranged to the memory of storage computer executable instructions, the executable instruction makes the place when executed
Manage the operation that device executes any of the above-described the method.
Compared with prior art, present invention could apply to chassis station personnel in automotive vehicle annual test to detect, and pass through
Facial image in the image of the car test station is compared with staff's face template image, obtains the first comparison knot
Fruit；Clothes area image in the image of the car test station is compared with staff's clothes template image, obtains
Two comparison results；By the mutual alignment relation and template human body between the human body key point position in the image of the car test station
Mutual alignment relation between key point position is compared, and obtains third comparison result；It is compared based on first, second, and third
As a result judge that the staff of car test station detects whether qualification, improve the accuracy rate of personnel's detection, and realize staff
Identification, dressing identification and compulsory exercise identification, while the whole-process automatic detection of review process, have not only saved manpower, but also
It ensure that the disclosure, just of detection work.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, of the invention other
Feature, objects and advantages will become more apparent upon:
Fig. 1 shows the flow chart of staff's detection method of the car test station of one embodiment of the invention.
The same or similar appended drawing reference represents the same or similar component in attached drawing.
Specific embodiment
Present invention is further described in detail with reference to the accompanying drawing.
In a typical configuration of this application, terminal, the equipment of service network and trusted party include one or more
Processor (CPU), input/output interface, network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flashRAM).Memory is showing for computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or
Any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, computer
Readable medium does not include non-temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
The present invention provides a kind of staff's detection method of car test station, which comprises
Step S1 obtains the image of car test station；
Here, the image of the car test station can be the image of the chassis station of car test, chassis station can be acquired
Image, and upload onto the server, when needing to detect, the image of chassis station is obtained from server；
Step S2 compares the facial image in the image of the car test station with staff's face template image
It is right, obtain the first comparison result；
Step S3 carries out the clothes area image in the image of the car test station with staff's clothes template image
It compares, obtains the second comparison result；
Step S4, by the mutual alignment relation and template between the human body key point position in the image of the car test station
Mutual alignment relation between human body key point position is compared, and obtains third comparison result；
Step S5 judges that the staff of car test station detects whether qualification based on the first, second, and third comparison result.
Here, present invention could apply to chassis station personnel in automotive vehicle annual test to detect, by by the car test
Facial image in the image of station is compared with staff's face template image, obtains the first comparison result；It will be described
Clothes area image in the image of car test station is compared with staff's clothes template image, obtains the second comparison knot
Fruit；By the mutual alignment relation and template human body key point between the human body key point position in the image of the car test station
Mutual alignment relation between setting is compared, and obtains third comparison result；Judged based on the first, second, and third comparison result
The staff of car test station detects whether qualification, improve personnel detection accuracy rate, and realize staff's identification,
Dressing identification and compulsory exercise identification, while the whole-process automatic detection of review process, have both saved manpower, have in turn ensured detection
It is the disclosure of work, just.
In one embodiment of staff's detection method of car test station of the invention, step S2, by the car test station
Facial image in image is compared with staff's face template image, obtains the first comparison result, comprising:
Step S21, using personnel's detection model based on deep learning network, judge be in the image of the car test station
No there are personnel area images, and personnel area image, then mark the corresponding personnel area of the image of the car test station if it does not exist
Will position is marked；Personnel area image if it exists then uses personnel's disaggregated model based on deep learning network, described in judgement
It whether there is personnel in the image of personnel area, if it does not exist personnel, then to the image of the car test station corresponding personnel identifier position
It is marked；
Step S22, personnel, obtain the personnel area image if it exists, are examined using the face based on deep learning network
Model is surveyed, detects and whether there is human face region image in the personnel area image, if it does not exist human face region image, then to this
The corresponding face flag bit of the image of car test station is marked；Face if it exists then obtains the human face region image, by institute
It states human face region image to be compared with staff's face template image, obtains the first comparison result.
Here, as shown in Figure 1, using in personnel's detection model detection chassis station image based on deep learning network
Personnel, and judge whether to obtain personnel area image,
If failing to get personnel area image, the corresponding personnel area flag bit of the image of the car test station is carried out
Labeled as 0, process terminates；
If personnel area image can be got, the people is judged in conjunction with personnel's disaggregated model based on deep learning network
Member area image in personnel whether necessary being,
If it does not exist, then the image of the car test station corresponding personnel identifier position is carried out being designated as 0, process terminates；
If it exists, downstream can be entered；
The Face datection model inspection face based on deep learning network is used in the personnel area image, judges people
Face area image whether there is,
Human face region image if it does not exist, then the corresponding face flag bit of the image of the car test station is marked is 0,
Into downstream；
Human face region image if it exists obtains human face region image, and by the face area image, with the work in database
Make personnel's face template image and carry out face alignment, judge whether the personnel are staff, records correlated results, for example,
If Face datection is qualified, by corresponding face, whether mark of conformity position is labeled as 1, subsequently into downstream；
If Face datection is unqualified, by corresponding face, whether mark of conformity position is labeled as 0, subsequently into lower one stream
Journey.
In one embodiment of staff's detection method of car test station of the invention, in step S21, using based on depth
The personnel's detection model for practising network judges in the image of the car test station with the presence or absence of before the image of personnel area, further includes:
Step S211 obtains the image of the sample car test station under different illumination, different shooting angles；
Step S212 marks personnel position in the image of the sample car test station using rectangle frame, and is labeled as
Personnel, the sample image after being marked；
Step S213 detects deep neural network model using the sample image training objective after the label, to obtain
Personnel's detection model based on deep learning network.
In one embodiment of staff's detection method of car test station of the invention, in step S21, using based on depth
The personnel's disaggregated model for practising network judges in the personnel area image with the presence or absence of before personnel, further includes:
S214. using personnel's detection model based on deep learning network, the figure of the sample car test station is obtained
The sample personnel area image of different location as in；
S215. classify to sample personnel area image, be divided into two class of someone's image and unmanned image；
S216. someone's image and unmanned image training objective depth of assortment neural network model are used, is based on
Personnel's disaggregated model of deep learning network.
In one embodiment of staff's detection method of car test station of the invention, in step S22, using based on depth
The Face datection model for practising network detects in the personnel area image with the presence or absence of before human face region image, further includes:
S221. different sample personnel area images is obtained；
S222. the human face region image position in the image of the sample personnel area is marked using rectangle frame, and marked
It is denoted as face；
S223. deep neural network model is detected using the sample personnel area image training objective after label, to obtain
Face datection model based on deep learning network.
In one embodiment of staff's detection method of car test station of the invention, step S3, by the car test station
Clothes area image in image is compared with staff's clothes template image, obtains the second comparison result, comprising:
Step S31, using the clothes detection model based on deep learning network, judge be in the image of the car test station
It is no there are clothes area image, clothes area image if it does not exist, then to the corresponding clothing logos position of the image of the car test station
It is marked；
Step S32, clothes area image, then obtain the clothes area image if it exists, by the clothes area image
It is compared with staff's clothes template image, obtains the second comparison result.
Here, as shown in Figure 1, detecting people using the clothes detection model based on deep learning network in personnel area image
Member's clothes, judge that clothes administrative division map whether there is,
Clothes area image if it does not exist, it is 0 that the corresponding clothing logos position of the image of the car test station, which is marked, into
Enter downstream；
Clothes area image if it exists, obtains clothes area image, and by the work in the clothes area image and database
Make personnel's clothes template figure to be compared, judge whether personnel's clothes meet regulation, records correlated results, for example,
If the detection of personnel's clothes is qualified, by corresponding clothes, whether mark of conformity position is labeled as 1, subsequently into lower one stream
Journey；
If the detection of personnel's clothes is unqualified, by corresponding clothes, whether mark of conformity position is labeled as 0, subsequently into next
Process.
In one embodiment of staff's detection method of car test station of the invention, step S31, using based on deep learning
The clothes detection model of network judges in the image of the car test station with the presence or absence of before clothes area image, further includes:
S311. different sample personnel area images is obtained；
S312. it using clothes area image position in rectangle frame label sample personnel area figure, and is labeled as
Clothes；
S313. depth neural network model is detected using the personnel area image training objective after label, to be based on
The clothes detection model of deep learning network.
In one embodiment of staff's detection method of car test station of the invention, step S4, by the car test station
It closes mutual alignment between the mutual alignment relation between human body key point position in image and template human body key point position
System is compared, and obtains third comparison result, comprising:
Step S41, using the human body critical point detection network model based on deep learning, in the personnel area image
Obtain human body key point position, if human body key point position has not been obtained, pass corresponding to the image of the car test station
Key point flag bit is marked；
Step S42, if getting human body key point position, the mutual position for the human body key point position that will acquire
The mutual alignment relation set between relationship and template human body key point position is compared, and obtains third comparison result.
Here, as shown in Figure 1, using the human body critical point detection network model based on deep learning in personnel area image
The key points such as testing staff's head, shoulder, elbow, wrist, chest, and these key point positions are obtained,
If failing to obtain these key point positions, the corresponding key point flag bit of the image of the car test station is marked
It is 0, into downstream；
If these key point positions can be obtained, using the mutual positional relationship of these key points, judge staff's
Whether movement meets regulation, records correlated results, for example,
If it is qualified that key point position is surveyed, by corresponding key point position, whether mark of conformity position is labeled as 1, subsequently into
Downstream；
If key point position detection is unqualified, by corresponding key point position, whether mark of conformity position is labeled as 0, then
Into downstream.
In one embodiment of staff's detection method of car test station of the invention, step S41, using based on deep learning
Human body critical point detection network model, in the personnel area image obtain human body key point position before, further includes:
S411. different sample personnel area images is obtained；
S412. label includes the crown, neck, left shoulder, right shoulder, left hand elbow, the right hand in the image of each sample personnel area
The key points coordinate such as elbow, left finesse, right finesse, thoracic cavity, pelvis；
S413. crucial Point Target Detection deep neural network model is trained using the sample personnel area image after label,
To obtain the human body critical point detection network model based on deep learning.
In one embodiment of staff's detection method of car test station of the invention, step S5, based on first, second and the
Three comparison results judge that the staff of car test station detects whether qualification, comprising:
Based on the personnel area flag bit, personnel identifier position, face flag bit, clothing logos position and described first,
Two and third comparison result, judge that the staff of car test station detects whether qualification.
Here, can be according to the personnel area flag bit, personnel identifier position, face flag bit, clothing logos position and institute
State the corresponding face of the first, second, and third comparison result whether mark of conformity position, clothes whether mark of conformity position, key
Point position whether mark of conformity position, judge that the staff of car test station detects whether qualification.For example, can be to whole process
As a result it is analyzed, if all flag bits of final entry are 1, illustrates that the image detection of car test station is qualified, if final entry
Having at least one in flag bit is 0, illustrates that the image detection of car test station is unqualified, which flag bit can be labeled as 0 according to,
Output detects unqualified reason and associated picture.
As shown in Figure 1, in one specific embodiment of staff's detection method of car test station of the invention, the method packet
It includes:
S101. the image of chassis station is obtained from server；
S102. using the personnel in the target detection model inspection chassis station image based on deep learning, and judgement is
No acquisition personnel area image, if failing to get personnel area image, personnel area corresponding to the image of the car test station
It is 0 that domain flag bit, which is marked, and process terminates；If personnel area image can be got, in conjunction with the target based on deep learning
Disaggregated model judge the personnel in the personnel area image whether necessary being, if it does not exist, then to the figure of the car test station
As corresponding personnel identifier position carries out being designated as 0, process terminates；If it exists, downstream can be entered；
S103. face is detected using the target detection network model based on deep learning in the personnel area image,
Judge that human face region image whether there is, if it does not exist human face region image, then to the corresponding face of the image of the car test station
It is 0 that flag bit, which is marked, into downstream；Human face region image if it exists obtains human face region image, and by the face
Staff's face template image in area image, with database carries out face alignment, judges whether the personnel are work people
Member records correlated results, for example, by corresponding face, whether mark of conformity position is labeled as 1, if people if Face datection is qualified
Face detection is unqualified, then by corresponding face, whether mark of conformity position is labeled as 0, subsequently into downstream；
S104. target detection network model testing staff's clothes based on deep learning are used in personnel area image, sentenced
Disconnected clothes administrative division map whether there is, if it does not exist clothes area image, clothing logos position corresponding to the image of the car test station
Being marked is 0, into downstream；Clothes area image if it exists obtains clothes area image, and by the clothes administrative division map
As being compared with staff's clothes template figure in database, judges whether personnel's clothes meet regulation, record related knot
Fruit, for example, whether mark of conformity position is labeled as 1 by corresponding clothes if the detection of personnel's clothes is qualified, if personnel's clothes detect
Unqualified, then by corresponding clothes, whether mark of conformity position is labeled as 0, subsequently into downstream；
S105. the human body critical point detection network model testing staff based on deep learning is used in personnel area image
The key points such as head, shoulder, elbow, wrist, chest, and these key point positions are obtained, if failing to obtain these key point positions, to the car test
It is 0 that the corresponding key point flag bit of the image of station, which is marked, into downstream；If these key points can be obtained
It sets, using the mutual positional relationship of these key points, judges whether the movement of staff meets regulation, record correlated results,
If it is qualified that key point position is surveyed, by corresponding key point position, whether mark of conformity position is labeled as 1, if key point position detection
Unqualified, then by corresponding key point position, whether mark of conformity position is labeled as 0, subsequently into downstream；
S106. the result of whole process is analyzed, if all flag bits of final entry are 1, illustrates chassis station
Photo is qualified, if having at least one in final entry flag bit is 0, illustrates that chassis station photo is unqualified, is gone out according to flag bit 0
Existing position, output detect unqualified reason and associated picture.
According to another aspect of the present invention, a kind of staff's detection device of car test station is additionally provided, the equipment packet
It includes:
Acquisition device, for obtaining the image of car test station；
First comparison device, for by the image of the car test station facial image and staff's face template figure
As being compared, the first comparison result is obtained；
Second comparison device, for by the image of the car test station clothes area image and staff's clothes mould
Domain picture is compared, and obtains the second comparison result；
Third comparison device, for by the mutual alignment between the human body key point position in the image of the car test station
Mutual alignment relation between relationship and template human body key point position is compared, and obtains third comparison result；
Judgment means, for judged based on the first, second, and third comparison result car test station staff detect be
No qualification.
According to another aspect of the present invention, a kind of equipment based on calculating is also provided characterized by comprising
Processor；And
It is arranged to the memory of storage computer executable instructions, the executable instruction makes the place when executed
Manage the operation that device executes any of the above-described the method.
The detailed content of each equipment and storage medium embodiment of the invention, for details, reference can be made to the correspondences of each method embodiment
Part, here, repeating no more.
Obviously, those skilled in the art can carry out various modification and variations without departing from the essence of the application to the application
Mind and range.In this way, if these modifications and variations of the application belong to the range of the claim of this application and its equivalent technologies
Within, then the application is also intended to include these modifications and variations.
It should be noted that the present invention can be carried out in the assembly of software and/or software and hardware, for example, can adopt
With specific integrated circuit (ASIC), general purpose computer or any other realized similar to hardware device.In one embodiment
In, software program of the invention can be executed to implement the above steps or functions by processor.Similarly, of the invention
Software program (including relevant data structure) can be stored in computer readable recording medium, for example, RAM memory,
Magnetic or optical driver or floppy disc and similar devices.In addition, some of the steps or functions of the present invention may be implemented in hardware, example
Such as, as the circuit cooperated with processor thereby executing each step or function.
In addition, a part of the invention can be applied to computer program product, such as computer program instructions, when its quilt
When computer executes, by the operation of the computer, it can call or provide according to the method for the present invention and/or technical solution.
And the program instruction of method of the invention is called, it is possibly stored in fixed or moveable recording medium, and/or pass through
Broadcast or the data flow in other signal-bearing mediums and transmitted, and/or be stored according to described program instruction operation
In the working storage of computer equipment.Here, according to one embodiment of present invention including a device, which includes using
Memory in storage computer program instructions and processor for executing program instructions, wherein when the computer program refers to
When enabling by processor execution, method and/or skill of the device operation based on aforementioned multiple embodiments according to the present invention are triggered
Art scheme.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included in the present invention.Any reference signs in the claims should not be construed as limiting the involved claims.This
Outside, it is clear that one word of " comprising " does not exclude other units or steps, and odd number is not excluded for plural number.That states in device claim is multiple
Unit or device can also be implemented through software or hardware by a unit or device.The first, the second equal words are used to table
Show title, and does not indicate any particular order.
Claims (10)
1. a kind of staff's detection method of car test station, which is characterized in that this method comprises:
Obtain the image of car test station；
Facial image in the image of the car test station is compared with staff's face template image, obtains the first ratio
To result；
Clothes area image in the image of the car test station is compared with staff's clothes template image, obtains
Two comparison results；
By the mutual alignment relation and template human body key point between the human body key point position in the image of the car test station
Mutual alignment relation between position is compared, and obtains third comparison result；
Judge that the staff of car test station detects whether qualification based on the first, second, and third comparison result.
2. the method according to claim 1, wherein by facial image and work in the image of the car test station
Make personnel's face template image to be compared, obtain the first comparison result, comprising:
Using personnel's detection model based on deep learning network, judge in the image of the car test station with the presence or absence of personnel area
Area image, personnel area image, then be marked the corresponding personnel area flag bit of the image of the car test station if it does not exist；
Personnel area image if it exists then uses personnel's disaggregated model based on deep learning network, judges the personnel area image
In whether there is personnel, personnel if it does not exist are then marked the image of the car test station corresponding personnel identifier position；
Personnel if it exists obtain the personnel area image, using the Face datection model based on deep learning network, detect institute
It states and whether there is human face region image in the image of personnel area, if it does not exist human face region image, then to the figure of the car test station
As corresponding face flag bit is marked；Face if it exists then obtains the human face region image, by the human face region figure
As being compared with staff's face template image, the first comparison result is obtained.
3. according to the method described in claim 2, it is characterized in that, using personnel's detection model based on deep learning network,
Judge in the image of the car test station with the presence or absence of before the image of personnel area, further includes:
Obtain the image of the sample car test station under different illumination, different shooting angles；
Personnel position in the image of the sample car test station is marked using rectangle frame, and is labeled as personnel, is marked
Sample image afterwards；
Deep neural network model is detected using the sample image training objective after the label, is based on deep learning net to obtain
Personnel's detection model of network.
4. according to the method in claim 2 or 3, which is characterized in that using personnel's classification mould based on deep learning network
Type judges in the personnel area image with the presence or absence of before personnel, further includes:
Using personnel's detection model based on deep learning network, different positions in the image of the sample car test station are obtained
The sample personnel area image set；
Classify to sample personnel area image, is divided into two class of someone's image and unmanned image；
Using someone's image and unmanned image training objective depth of assortment neural network model, obtains and be based on deep learning net
Personnel's disaggregated model of network.
5. according to the method described in claim 2, it is characterized in that, using the Face datection model based on deep learning network,
It detects in the personnel area image with the presence or absence of before human face region image, further includes:
Obtain different sample personnel area images；
Human face region image position in the image of the sample personnel area is marked using rectangle frame, and is labeled as face；
Deep neural network model is detected using the sample personnel area image training objective after label, is based on depth to obtain
Practise the Face datection model of network.
6. the method according to claim 1, wherein by the clothes area image in the image of the car test station
It is compared with staff's clothes template image, obtains the second comparison result, comprising:
Using the clothes detection model based on deep learning network, judge in the image of the car test station with the presence or absence of clothes area
Area image, clothes area image, then be marked the corresponding clothing logos position of the image of the car test station if it does not exist；
Clothes area image if it exists then obtains the clothes area image, and the clothes area image and staff are taken
Dress template image is compared, and obtains the second comparison result.
7. according to the method described in claim 6, it is characterized in that, using the clothes detection model based on deep learning network,
Judge in the image of the car test station with the presence or absence of before clothes area image, further includes:
Obtain different sample personnel area images；
Using clothes area image position in rectangle frame label sample personnel area figure, and it is labeled as clothes；
Depth neural network model is detected using the personnel area image training objective after label, is based on deep learning net to obtain
The clothes detection model of network.
8. the method according to claim 1, wherein by the human body key point in the image of the car test station
Mutual alignment relation between setting is compared with the mutual alignment relation between template human body key point position, obtains third ratio
To result, comprising:
Using the human body critical point detection network model based on deep learning, it is crucial that human body is obtained in the personnel area image
Point position, if human body key point position has not been obtained, to the corresponding key point flag bit of the image of the car test station into
Line flag；
If getting human body key point position, the mutual alignment relation and template of the human body key point position that will acquire
Mutual alignment relation between human body key point position is compared, and obtains third comparison result.
9. according to the method described in claim 8, it is characterized in that, using the human body critical point detection network based on deep learning
Model, in the personnel area image before acquisition human body key point position, further includes:
Obtain different sample personnel area images；
In the image of each sample personnel area label include the crown, neck, left shoulder, right shoulder, left hand elbow, right hand elbow, left finesse,
The key points coordinate such as right finesse, thoracic cavity, pelvis；
Using the crucial Point Target Detection deep neural network model of sample personnel area image training after label, to be based on
The human body critical point detection network model of deep learning.
10. according to the method described in claim 8, it is characterized in that, judging car test based on the first, second, and third comparison result
The staff of station detects whether qualification, comprising:
Based on the personnel area flag bit, personnel identifier position, face flag bit, clothing logos position and first, second He
Third comparison result judges that the staff of car test station detects whether qualification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910576535.6A CN110309768B (en) | 2019-06-28 | 2019-06-28 | Method and equipment for detecting staff at vehicle inspection station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910576535.6A CN110309768B (en) | 2019-06-28 | 2019-06-28 | Method and equipment for detecting staff at vehicle inspection station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110309768A true CN110309768A (en) | 2019-10-08 |
| CN110309768B CN110309768B (en) | 2020-11-20 |
Family
ID=68078640
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910576535.6A Expired - Fee Related CN110309768B (en) | 2019-06-28 | 2019-06-28 | Method and equipment for detecting staff at vehicle inspection station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110309768B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717449A (en) * | 2019-10-09 | 2020-01-21 | 上海眼控科技股份有限公司 | Vehicle annual inspection personnel behavior detection method and device and computer equipment |
| CN110852212A (en) * | 2019-10-29 | 2020-02-28 | 上海眼控科技股份有限公司 | Method and device for checking operation object in vehicle detection |
| CN111652046A (en) * | 2020-04-17 | 2020-09-11 | 济南浪潮高新科技投资发展有限公司 | Safe wearing detection method, equipment and system based on deep learning |
| CN112560817A (en) * | 2021-02-22 | 2021-03-26 | 西南交通大学 | Human body action recognition method and device, electronic equipment and storage medium |
| CN112907199A (en) * | 2021-01-22 | 2021-06-04 | 陕西交通电子工程科技有限公司 | Intelligent management system for assisting vehicle examination |
| CN114882597A (en) * | 2022-07-11 | 2022-08-09 | 浙江大华技术股份有限公司 | Target behavior identification method and device and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170032182A1 (en) * | 2015-03-26 | 2017-02-02 | vCognition Technologies Inc. | System for adaptive real-time facial recognition using fixed video and still cameras |
| CN107590807A (en) * | 2017-09-29 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Method and apparatus for detection image quality |
| CN107920223A (en) * | 2016-10-08 | 2018-04-17 | 杭州海康威视数字技术股份有限公司 | A kind of object behavior detection method and device |
| CN108174165A (en) * | 2018-01-17 | 2018-06-15 | 重庆览辉信息技术有限公司 | Electric power safety operation and O＆M intelligent monitoring system and method |
| CN108229855A (en) * | 2018-02-06 | 2018-06-29 | 上海小蚁科技有限公司 | Method for monitoring service quality and device, computer readable storage medium, terminal |
| CN109146322A (en) * | 2018-09-12 | 2019-01-04 | 深圳市商汤科技有限公司 | Monitoring method and device and system, electronic equipment and storage medium |
| CN109299683A (en) * | 2018-09-13 | 2019-02-01 | 嘉应学院 | A kind of security protection assessment system based on recognition of face and behavior big data |
-
2019
- 2019-06-28 CN CN201910576535.6A patent/CN110309768B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170032182A1 (en) * | 2015-03-26 | 2017-02-02 | vCognition Technologies Inc. | System for adaptive real-time facial recognition using fixed video and still cameras |
| CN107920223A (en) * | 2016-10-08 | 2018-04-17 | 杭州海康威视数字技术股份有限公司 | A kind of object behavior detection method and device |
| CN107590807A (en) * | 2017-09-29 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Method and apparatus for detection image quality |
| CN108174165A (en) * | 2018-01-17 | 2018-06-15 | 重庆览辉信息技术有限公司 | Electric power safety operation and O＆M intelligent monitoring system and method |
| CN108229855A (en) * | 2018-02-06 | 2018-06-29 | 上海小蚁科技有限公司 | Method for monitoring service quality and device, computer readable storage medium, terminal |
| CN109146322A (en) * | 2018-09-12 | 2019-01-04 | 深圳市商汤科技有限公司 | Monitoring method and device and system, electronic equipment and storage medium |
| CN109299683A (en) * | 2018-09-13 | 2019-02-01 | 嘉应学院 | A kind of security protection assessment system based on recognition of face and behavior big data |
Non-Patent Citations (2)
| Title |
|---|
| COLIN LEA 等: "Learning Convolutional Action Primitives for Fine-grained Action Recognition", 《IEEE》 * |
| 李长征 等: "对机动车检验实行智能监管的设想", 《道路交通管理》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717449A (en) * | 2019-10-09 | 2020-01-21 | 上海眼控科技股份有限公司 | Vehicle annual inspection personnel behavior detection method and device and computer equipment |
| CN110852212A (en) * | 2019-10-29 | 2020-02-28 | 上海眼控科技股份有限公司 | Method and device for checking operation object in vehicle detection |
| CN111652046A (en) * | 2020-04-17 | 2020-09-11 | 济南浪潮高新科技投资发展有限公司 | Safe wearing detection method, equipment and system based on deep learning |
| CN112907199A (en) * | 2021-01-22 | 2021-06-04 | 陕西交通电子工程科技有限公司 | Intelligent management system for assisting vehicle examination |
| CN112907199B (en) * | 2021-01-22 | 2024-02-02 | 陕西交通电子工程科技有限公司 | Intelligent management system for assisting vehicle inspection |
| CN112560817A (en) * | 2021-02-22 | 2021-03-26 | 西南交通大学 | Human body action recognition method and device, electronic equipment and storage medium |
| CN114882597A (en) * | 2022-07-11 | 2022-08-09 | 浙江大华技术股份有限公司 | Target behavior identification method and device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110309768B (en) | 2020-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110309768A (en) | The staff's detection method and equipment of car test station | |
| WO2021047232A1 (en) | Interaction behavior recognition method, apparatus, computer device, and storage medium | |
| CN104572804B (en) | A kind of method and its system of video object retrieval | |
| CN105574550B (en) | A kind of vehicle identification method and device | |
| CN104318225B (en) | Detection method of license plate and device | |
| CN107609493A (en) | Optimize the method and device of face picture Environmental Evaluation Model | |
| CN110490135A (en) | Vehicle annual test method, apparatus, computer equipment and storage medium | |
| CN109934182A (en) | Object behavior analysis method, device, electronic equipment and computer storage medium | |
| CN110348360A (en) | A kind of examining report recognition methods and equipment | |
| CN110378258B (en) | Image-based vehicle seat information detection method and device | |
| CN100361136C (en) | Evaluation method and system for face detection system | |
| CN112001219B (en) | Multi-angle multi-face recognition attendance checking method and system | |
| US20220207266A1 (en) | Methods, devices, electronic apparatuses and storage media of image processing | |
| CN108537218A (en) | A kind of identifying processing method and device of answering card | |
| CN104077568A (en) | High-accuracy driver behavior recognition and monitoring method and system | |
| CN109948616A (en) | Image detecting method, device, electronic equipment and computer readable storage medium | |
| CN107292318A (en) | Image significance object detection method based on center dark channel prior information | |
| CN109657622A (en) | A kind of detection of traffic lights and recognition methods, device and equipment | |
| CN111126393A (en) | Vehicle appearance refitting judgment method and device, computer equipment and storage medium | |
| CN110288006A (en) | A kind of license plate number automatic Verification method and system | |
| CN107909009A (en) | A kind of obstacle detection method and device based on road surface study | |
| CN104463240A (en) | Method and device for controlling list interface | |
| CN106156713A (en) | A kind of image processing method automatically monitored for examination hall behavior | |
| CN114565976A (en) | Training intelligent test method and device | |
| CN109118512A (en) | A kind of classroom based on machine vision is come to work late and leave early detection method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Staff detection method and equipment of vehicle inspection station Effective date of registration: 20220211 Granted publication date: 20201120 Pledgee: Shanghai Bianwei Network Technology Co.,Ltd. Pledgor: SHANGHAI EYE CONTROL TECHNOLOGY Co.,Ltd. Registration number: Y2022310000023 |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20201120 |