CN115620385A - Multivariate data-based security check worker attention detection method and system - Google Patents
Multivariate data-based security check worker attention detection method and system Download PDFInfo
- Publication number
- CN115620385A CN115620385A CN202211381852.0A CN202211381852A CN115620385A CN 115620385 A CN115620385 A CN 115620385A CN 202211381852 A CN202211381852 A CN 202211381852A CN 115620385 A CN115620385 A CN 115620385A
- Authority
- CN
- China
- Prior art keywords
- security
- detected
- attention
- ray
- article
- 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
- 238000001514 detection method Methods 0.000 title claims description 69
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000003384 imaging method Methods 0.000 claims description 30
- 238000007689 inspection Methods 0.000 claims description 18
- 238000013528 artificial neural network Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 2
- 230000000306 recurrent effect Effects 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 229910052739 hydrogen Inorganic materials 0.000 claims 1
- 239000001257 hydrogen Substances 0.000 claims 1
- 125000004435 hydrogen atom Chemical class [H]* 0.000 claims 1
- 238000013135 deep learning Methods 0.000 abstract description 2
- 238000011156 evaluation Methods 0.000 abstract description 2
- 230000008569 process Effects 0.000 abstract description 2
- 210000003128 head Anatomy 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 206010041349 Somnolence Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000004424 eye movement Effects 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Images
Classifications
-
- 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
- G06V40/18—Eye characteristics, e.g. of the iris
- G06V40/19—Sensors therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- 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
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Abstract
The invention discloses a method and a system for detecting the attention of security check workers based on multivariate data, which quantitatively evaluate the working state and the working quality of the security check workers by extracting the characteristics of a fixation point of the security check workers, the pixel speed of an X-ray machine, the intelligent article identification result of an X-ray image and the like and utilizing a model based on deep learning. The invention can intelligently identify the X-ray image and quantitatively evaluate the working state and the working quality of the security inspector, provides data support for the performance evaluation of the security inspector, enhances the control on the working process of the security inspector, reduces the problem of omission of the security inspector and lowers the management cost of a human company.
Description
Technical Field
The invention relates to the field of security detection, in particular to a method and a system for detecting attention of security check workers based on multivariate data.
Background
At present, passengers in public transportation places such as airports, high-speed rails, subways and the like need to put personal packages into an X-ray machine for security inspection before entering a station. The security inspector discovers potential dangerous goods in the passenger package through the X-ray machine and eliminates potential safety hazards in the station. Due to the nature of the security check work, it is difficult to quantitatively evaluate the working state and the working quality of the security checker.
The current technical means mainly utilizes an image analysis technology or an image recognition technology to analyze eye movement data or facial feature data to judge the fatigue state, working emotion or attention direction of a security inspector, and only can distinguish whether the security inspector is sleepy or not at present. It is impossible to judge whether the security inspector is working seriously or in a vague state, and whether the security inspector is missing the inspection currently. Meanwhile, due to different professional technical levels of security inspectors, different security inspectors judge that the watching time of the same article is different. If only the direction of the gazing point and the position of the gazing point are relied on, the work of the security inspector cannot be accurately judged.
In contrast, the invention patent CN 110705500A discloses a method and a system for detecting attention of a worker working image based on deep learning, which supervises and records the attention of the worker through "a head posture angle, the two eye gazing directions, and the eye opening and closing states; when the human eyes are in an eye closing state, the head posture angle exceeds a specified threshold value, and the two-eye gazing direction exceeds a specified range, any one or any combination of the three conditions can be triggered to remind. However, the above method merely detects whether the eyes of a person are focused on the display screen (because the display screen does not move, the head posture angle is a fixed range), and cannot judge whether the attention of the worker is distracted when the worker watches the display screen, and whether contraband which should be watched for inspection is missed, which results in that the judgment on whether the attention of the worker is focused is not accurate enough.
In short, the current technical means cannot accurately and effectively control the working process of the security inspector and cannot accurately judge the working quality of the security inspector, so that a method for controlling the working state of the security inspector, judging the working quality of the security inspector and assisting the security inspector to check forbidden band articles is urgently needed.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method and a system for detecting the attention of security personnel based on multivariate data in order to realize the detection of the working state and the working quality of the security personnel.
In order to realize the purpose, the technical scheme of the invention is as follows:
a method for detecting attention of security check workers based on multivariate data comprises the following steps:
collecting fixation point data when different security check workers check an X-ray image and corresponding position data of an object to be detected of the X-ray image, performing data alignment on the fixation point data of the workers when the X-ray image is checked, forming a data set according to the fidelity of the security check workers when the X-ray image is checked, inputting the data set into a neural network, and training to obtain an attention detection model for judging whether the workers pay attention to the object to be detected;
secondly, placing the article to be detected into an inspection channel of an X-ray machine for detection and identification, displaying a detection image on a display screen of the X-ray machine, and obtaining the position coordinates of the article to be detected on the display screen of the X-ray machine in real time;
step three, the fixation point of the safety inspector sight is obtained by an eye tracker, and the data is aligned and converted into the fixation coordinate point on the X-ray machine video picture (x v ，y v ）；
Step four, calculating the pixel size of an image on a display screen of an X-ray machine, determining the pixel position and the imaging duration of the to-be-detected article in a video stream appearing on the display screen according to the pixel moving speed, then collecting the watching coordinate point of a security inspector in the imaging duration of the to-be-detected article, and calculating whether the watching coordinate point is in the pixel position of the to-be-detected article to form a result sequence; inputting the result sequence within the imaging duration into an attention detection model for detection to obtain the probability of whether a security inspector notices an article to be detected;
and step five, when the probability that the security check worker notices the article to be detected is lower than a preset threshold value, judging that the security check worker is in an inattentive state, and alarming and reminding.
Further improvement, the method also comprises a sixth step of judging the current working state of a security inspector: get the current and past continuationsWDetecting the attention of the secondary X-ray package image to obtain a detection result set, and determining whether the probability value in the detection result set is greater than a threshold valueNNumber of times ofGIn [0,1 ]]The different working attitude intervals are divided into corresponding different working attitude degrees, the current working state of the corresponding user is judged according to the working attitude interval in which the value of G/W is positioned, N belongs to [0,1 ]]。
In a further refinement, the neural network is a recurrent neural network.
In a further improvement, in the first step, the gazing point data includes: the fixation point coordinate of the staff on the X-ray machine display detected by the eye tracker (1:)x e ,y e ) The contraband position data of the X-ray image is the position of the X-ray image of the article to be detected in the video stream of the security check machine; wherein the content of the first and second substances,x e ,y e respectively an abscissa and an ordinate of a fixation point of security inspection work; establishing a plane coordinate system of the X-ray video stream of the security check machine on a plane where the display screen of the X-ray machine is located, wherein the plane coordinate system of the X-ray video stream of the security check machine takes the resolution of the display screen of the X-ray machine as a basic unit; the data alignment method is realized byh o ，w o ，h p ，w p To the fixation point coordinate
Performing coordinate transformation, and converting the fixation point coordinate into a plane coordinate system of the X-ray video stream of the security inspection machine;h o ，w o respectively representing the resolution of the X-axis and y-axis of the pictures of the X-ray video stream of the security inspection machine,h p ，w p respectively representing the x-axis and y-axis resolution of the current display.
In a further refinement, in step four, the pixel movement speedv pix Obtained by the following method:
passing standard test substance with known actual lengthThe inspection machine records the time from absence of the standard test object to complete imaging in the X-ray video stream of the security inspection machine; calculating the pixel length of the standard test object completing imaging in the X-ray video stream of the security inspection machine; the length of the pixel which finishes imaging is divided by the time from no imaging to complete imaging to obtain the corresponding pixel moving speedv pix 。
In a further improvement, in the fourth step, the method for forming the result sequence is as follows:
firstly, calculating the imaging time length of an article to be detected appearing in a video stream of a display screen
Then calculating the X-ray image and pixel position set of the article to be detected within the imaging duration
And image width and heightw，h）；
The abscissa representing the top left vertex of the X-ray image of the object to be inspected at the t-th time point,
a vertical coordinate representing the top left vertex of the X-ray image of the object to be detected at the t-th time point;
acquiring an attention point set of security check workers and aligning data to obtain a coordinate point set
；x t ，y t Respectively representing the abscissa and the ordinate of the attention point of the security check worker at the t-th time point;
judging the pixel area of the object to be detected linked with the point in the coordinate point set, judging whether each point in the coordinate point set falls in the X-ray wrapped image or not, and judging a result r i The following were used:
forming the judgment result set into a judgment result set
Judgment result set
Sending the data into an attention detection model for detection to obtain the probability that a user notices an article to be detected; r is i Indicating whether the attention point of the security check worker falls within the pixel area of the item to be detected at the ith time point,
the abscissa representing the top left vertex of the rectangular frame of the object to be inspected at the ith instant,
the ordinate representing the top left vertex of the rectangular frame of the object to be inspected at the ith instant,x i abscissa, y, representing the point of attention of the security worker at the ith time point i The ordinate representing the point of attention of the security check worker at the ith time point,wthe width of the rectangular frame of the object to be inspected is shown,hindicating the height of the rectangular frame of the article to be inspected.
In the second step, when the articles to be detected contain contraband, the contraband is marked on the display screen through a contraband detection frame; simultaneously obtaining the probability of whether the security inspector notices the contraband detection frame in the fourth step; in the second step, when the object to be detected does not contain contraband, displaying a small external rectangle on the periphery of the object to be detected in the display screen as a package detection frame; and in the fourth step, the probability of whether the security inspector notices the package detection box is obtained at the same time.
The method comprises the following steps that when the articles to be detected contain contraband, a minimum circumscribed rectangular frame is displayed outside the contraband to serve as a contraband detection frame, whether the probability that a security inspector notices the contraband detection frame is lower than a preset threshold value or not is calculated, and when the probability that the security inspector notices the articles to be detected is lower than the preset threshold value, the security inspector is judged to be in a state of inattention and is alarmed;
the labeling method of the contraband detection frame comprises the following steps:
s1, inputting X-ray images of various contraband articles and corresponding equivalent atomic number information into a target detection neural network for training to obtain a trained article detection model for identifying articles;
s2, placing the article to be detected into an inspection channel of an X-ray machine, acquiring bottom layer data of the X-ray machine, and extracting a complete package image through a package segmentation algorithm;
and S3, inputting the complete package image into a trained article detection model for identification, obtaining the article type of contraband, a contraband detection frame and confidence coefficient, and superposing the obtained information on the original video stream of the security check machine in real time for display.
And further improving, wherein the position data of the minimum circumscribed rectangle of the article to be detected is used as the position data of the article to be detected.
A security check worker attention detection system based on multivariate data is characterized by comprising an X-ray machine and computer equipment; the computer equipment is used for realizing the method for detecting the attention of the security check worker based on the multivariate data.
In summary, compared with the prior art, the technical scheme of the invention has the following advantages:
according to the invention, the eye tracker is used for acquiring the gaze point of the security inspector, the neural network is used for identifying contraband, the gaze point and the information of the contraband are input into the neural network model to judge the probability of the security inspector noticing the contraband, the current working quality and working state of the user are judged according to probability values of a certain number of times, and whether the attention of the security inspector is in a concentrated state can be more accurately judged.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic view of a point of regard within a contraband detection frame;
fig. 3 is a schematic diagram of the fixation point outside the contraband detection frame.
Detailed Description
In order to more clearly describe the technical scheme and advantages of the method of the present invention, the technical scheme and the working principle of the present invention are described in more detail with reference to the accompanying drawings and embodiments.
As shown in FIG. 1, a method for detecting the attention of security personnel based on multivariate data comprises the following steps:
the method comprises the steps of firstly, collecting gazing point data and X-ray image data when different users carefully check an X-ray image, aligning the gazing point data with an X-ray machine video image, inputting the gazing point data and the X-ray machine video image into a neural network in a combined mode, and training to obtain an attention detection model for judging whether the users pay attention to an article.
Inputting the X-ray image of the article and the corresponding equivalent atomic number information into a neural network for training to obtain a trained article detection model for identifying the article;
and step three, placing the article to be detected into an inspection channel of an X-ray machine, acquiring bottom layer data of the X-ray machine, and extracting a complete package image through a package segmentation algorithm.
And step four, inputting the parcel image into an article detection model for identification to obtain the article type, an article detection frame and confidence, and superposing the obtained information on the original video stream of the security inspection machine in real time for display.
Step five, the fixation point of the sight of the security inspector is obtained through an eye tracker, and data alignment is carried out to convert the fixation point into a fixation coordinate point on the video picture of the X-ray machine (x v ，y v ）。
Step six, calculating the pixel size of the parcel image, determining the pixel position and the imaging time of the parcel in the video stream according to the pixel moving speed, then acquiring the gazing coordinate point of a security inspector in the imaging time, and calculating whether the gazing coordinate point is in the pixel position of the parcel image; and inputting the result sequence in the imaging time period into an attention detection model for detection to obtain the probability of whether the package is noticed by a security inspector.
And seventhly, determining the working quality and the working state of the user according to the attention detection result of the user, the X-ray package image position information and the contraband detection frame information.
Data alignment
The data alignment steps described in the first and fifth steps are as follows:
1. acquiring the image resolution of the current security check equipmenth o ，w o ）；
2. Obtaining the resolution of the current display: (h p ，w p ）；
3. Acquiring coordinate points of the fixation points of the staff detected by the eye tracker (x e ,y e )；
4. The attention point is converted into a coordinate point (x, y) on the video screen of the security check machine.
Pixel moving speed calculation
Pixel moving speed as described in step six
The method comprises the following steps:
enabling a standard test object with a known actual length to pass through a security check machine, and recording the time from absence to complete imaging of the standard test object in an X-ray video stream of the security check machine; calculating the pixel length of the standard test object completing imaging in the X-ray video stream of the security inspection machine; the length of the pixel which finishes imaging is divided by the time from no imaging to complete imaging to obtain the corresponding pixel moving speedv pix
3. Attention detection
The attention detection method as described in step six, firstly calculating the imaging time
Then calculating the pixel position set of the X-ray package image and the contraband detection frame in the imaging time
And image width and heightw，h）；
Acquiring a set of attention points of a user and aligning data to obtain a set of coordinate points
；
Judging the pixel position of the point linkage X-ray package image or the contraband detection frame in the coordinate point set, judging whether each point falls in the X-ray package image or not, and judging the result
The following were used:
set the judgment result
And sending the X-ray package image into an attention detection model for detection to obtain the probability that a user notices the X-ray package image.
Evaluation of working quality and working state
The method for judging the working quality and the working state in the step seven comprises the following steps:
the method for judging the current attention condition of the security inspector comprises the following steps: get the
Detecting the attention of the detection frame containing contraband, and determining whether the concentrated probability value of the detection result is greater than the threshold value
Number of times of
Quantitatively determining the working quality of the user, wherein
And is provided with
。
The method for judging the current working state of the security inspector comprises the following steps: get the current and past continuations
The attention detection result set of the secondary X-ray package image is detected according to the fact that the probability value in the detection result set is larger than the threshold value
Number of times of
Determining the current working state of the user, wherein
And is
。
The above description is only a preferred embodiment of the present invention, and it should be noted that any person skilled in the art can use the above method to change the technical scheme of the present invention in various forms or modify the technical scheme of the present invention into an equivalent embodiment. Therefore, any simple modification or equivalent changes made according to the technical method of the present invention are within the scope of the protection of the method of the present invention, unless the scope of the technical solution of the present invention is exceeded.
Claims (10)
1. A method for detecting attention of security check workers based on multivariate data is characterized by comprising the following steps:
collecting fixation point data when different security check workers check an X-ray image and corresponding position data of an article to be detected of the X-ray image, carrying out data alignment on the fixation point data of the workers when the X-ray image is checked, forming a data set according to the degree of seriousness when the security check workers check the X-ray image, inputting the data set into a neural network, and training to obtain an attention detection model for judging whether a user pays attention to the article to be detected;
secondly, placing the article to be detected into a detection channel of an X-ray machine for detection and identification, displaying a detection image on a display screen of the X-ray machine, and obtaining the position coordinates of the article to be detected on the display screen of the X-ray machine in real time;
step three, the fixation point of the safety inspector sight is obtained by an eye tracker, and the data is aligned and converted into the fixation coordinate point on the X-ray machine video picture (x v ，y v ）；
Step four, calculating the pixel size of an image on a display screen of an X-ray machine, determining the pixel position and the imaging duration of the to-be-detected article in a video stream appearing on the display screen according to the pixel moving speed, then collecting the watching coordinate point of a security inspector in the imaging duration of the to-be-detected article, and calculating whether the watching coordinate point is in the pixel position of the to-be-detected article to form a result sequence; inputting the result sequence within the imaging duration into an attention detection model for detection to obtain the probability of whether a security inspector notices an article to be detected;
and step five, when the probability that the security inspector notices the article to be detected is lower than a preset threshold value, judging that the security inspector is in an inattentive state, and alarming and reminding.
2. The multivariate data-based method for detecting the attention of security personnel according to claim 1, further comprising the step six of judging the current working state of the security personnel: get the current and past continuanceWThe attention detection result set of the secondary X-ray package image is detected according to the fact that the probability value in the detection result set is larger than the threshold valueNNumber of times ofGIn [0,1 ]]The different working attitude intervals are divided into corresponding different working attitude degrees, the current working state of the corresponding user is judged according to the working attitude interval in which the value of G/W is positioned, N belongs to [0,1 ]]。
3. The multivariate data-based method for detecting attention of security personnel according to claim 1, wherein the neural network is a recurrent neural network.
4. The method for detecting the attention of the security worker based on the multivariate data as claimed in claim 1, wherein in the first step, the gazing point data comprises: the fixation point coordinate of the staff on the X-ray machine display detected by the eye tracker (1:)x e ,y e ) The contraband position data of the X-ray image is the position of the X-ray image of the article to be detected in the video stream of the security check machine; wherein the content of the first and second substances,x e ,y e respectively an abscissa and an ordinate of a fixation point of security inspection work; establishing a plane coordinate system of the X-ray video stream of the security check machine on a plane where the display screen of the X-ray machine is located, wherein the plane coordinate system of the X-ray video stream of the security check machine takes the resolution of the display screen of the X-ray machine as a basic unit; the data alignment method is realized byh o ，w o ，h p ，w p To the fixation point coordinate
Performing coordinate transformation, and converting the fixation point coordinate into a plane coordinate system of the X-ray video stream of the security inspection machine;h o ，w o respectively representing the resolution of the X-axis and y-axis of the pictures of the X-ray video stream of the security inspection machine,h p ，w p respectively representing the x-axis and y-axis resolution of the current display.
5. The method for detecting attention of security personnel based on multivariate data as defined in claim 1, wherein in the fourth step, the pixel moving speedv pix The method comprises the following steps:
enabling a standard test object with a known actual length to pass through a security check machine, and recording the time from absence to complete imaging of the standard test object in an X-ray video stream of the security check machine; x-ray video of security check machine for calculating standard test substanceThe length of the pixel in the stream at which imaging is completed; the length of the pixel completing imaging is divided by the time from no imaging to full imaging, and the corresponding pixel moving speed is calculatedv pix 。
6. The multivariate data-based method for detecting the attention of security personnel according to claim 1, wherein in the fourth step, the method for forming the result sequence comprises the following steps:
firstly, calculating the imaging time length of an article to be detected appearing in a video stream of a display screen
Then calculating the X-ray image and pixel position set of the article to be detected within the imaging duration
And image width and heightw，h）；
The abscissa representing the top left vertex of the X-ray image of the object to be inspected at the t-th time point,
the ordinate of the top left vertex of the X-ray image of the object to be detected at the t-th time point is represented;
acquiring an attention point set of security check workers and aligning data to obtain a coordinate point set
；x t ，y t Respectively representing the abscissa and the ordinate of the attention point of the security check worker at the t-th time point;
judging the pixel area of the object to be detected linked with the point in the coordinate point set, judging whether each point in the coordinate point set falls in the X-ray wrapped image or not, and judging a result r i The following:
forming a judgment result set by the judgment result set
Judgment result set
Sending the data to an attention detection model for detection to obtain the probability of a user paying attention to the object to be detected; r is a radical of hydrogen i Indicating whether the point of attention of the security worker at the ith time point falls within the pixel region of the item to be inspected,
the abscissa representing the upper left vertex of the rectangular frame of the article to be inspected at the ith instant,
the ordinate representing the top left vertex of the rectangular frame of the object to be inspected at the ith instant,x i abscissa, y, representing the point of attention of the security worker at the ith time point i The ordinate representing the point of attention of the security check worker at the ith time point,wthe width of the rectangular frame of the object to be inspected is shown,hindicating the height of the rectangular frame of the object to be inspected.
7. The method for detecting the attention of the security check worker based on the multivariate data as claimed in claim 1, wherein in the second step, when the articles to be detected contain contraband, the contraband is marked on the display screen through a contraband detection frame; simultaneously obtaining the probability of whether the security inspector notices the contraband detection frame in the fourth step; in the second step, when the object to be detected does not contain contraband, displaying a small external rectangle on the periphery of the object to be detected in the display screen as a package detection frame; and in the fourth step, the probability of whether the security inspector notices the package detection box is obtained simultaneously.
8. The method for detecting the attention of the security personnel based on the multivariate data as claimed in claim 7, wherein when the articles to be detected contain contraband, a minimum circumscribed rectangle frame is shown outside the contraband as a contraband detection frame, whether the probability that the security personnel notices the contraband detection frame is lower than a preset threshold value or not is calculated, and when the probability that the security personnel notices the articles to be detected is lower than the preset threshold value, the security personnel is judged to be in a state of inattention and alarm reminding is performed;
the labeling method of the contraband detection frame comprises the following steps:
s1, inputting X-ray images of various contraband articles and corresponding equivalent atomic number information into a target detection neural network for training to obtain a trained article detection model for identifying articles;
s2, placing the article to be detected into an inspection channel of an X-ray machine, acquiring bottom layer data of the X-ray machine, and extracting a complete package image through a package segmentation algorithm;
and S3, inputting the complete package image into a trained article detection model for identification, obtaining the article type of contraband, a contraband detection frame and confidence coefficient, and superposing the obtained information on the original video stream of the security check machine in real time for display.
9. The multivariate data-based security worker attention detection method as claimed in claim 1, wherein position data of a minimum circumscribed rectangle of the article to be detected is taken as the position data of the article to be detected.
10. A security check worker attention detection system based on multivariate data is characterized by comprising an X-ray machine and computer equipment; the computer equipment is used for realizing the multivariate data-based security personnel attention detection method as defined in any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211381852.0A CN115620385B (en) | 2022-11-07 | 2022-11-07 | Safety inspection staff attention detection method and system based on multivariate data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211381852.0A CN115620385B (en) | 2022-11-07 | 2022-11-07 | Safety inspection staff attention detection method and system based on multivariate data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115620385A true CN115620385A (en) | 2023-01-17 |
| CN115620385B CN115620385B (en) | 2023-07-28 |
Family
ID=84877880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211381852.0A Active CN115620385B (en) | 2022-11-07 | 2022-11-07 | Safety inspection staff attention detection method and system based on multivariate data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115620385B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456129A (en) * | 2010-10-26 | 2012-05-16 | 同方威视技术股份有限公司 | Image deviation-rectifying method and system for safety inspection |
| CN107798825A (en) * | 2017-10-26 | 2018-03-13 | 中控智慧科技股份有限公司 | Safety inspector's based reminding method and device |
| WO2018150024A1 (en) * | 2017-02-17 | 2018-08-23 | Cogisen S.R.L. | Method for image processing and video compression |
| CN110543857A (en) * | 2019-09-05 | 2019-12-06 | 安徽启新明智科技有限公司 | Contraband identification method, device and system based on image analysis and storage medium |
| AU2020101324A4 (en) * | 2018-05-07 | 2020-08-20 | Apple Inc. | Multi-participant live communication user interface |
| CN112949461A (en) * | 2021-02-26 | 2021-06-11 | 北京高途云集教育科技有限公司 | Learning state analysis method and device and electronic equipment |
| CN113807119A (en) * | 2020-05-29 | 2021-12-17 | 魔门塔(苏州)科技有限公司 | Method and device for detecting person fixation position |
| CN114120432A (en) * | 2021-11-17 | 2022-03-01 | 湖北大学 | Online learning attention tracking method based on sight estimation and application thereof |
-
2022
- 2022-11-07 CN CN202211381852.0A patent/CN115620385B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456129A (en) * | 2010-10-26 | 2012-05-16 | 同方威视技术股份有限公司 | Image deviation-rectifying method and system for safety inspection |
| WO2018150024A1 (en) * | 2017-02-17 | 2018-08-23 | Cogisen S.R.L. | Method for image processing and video compression |
| CN107798825A (en) * | 2017-10-26 | 2018-03-13 | 中控智慧科技股份有限公司 | Safety inspector's based reminding method and device |
| AU2020101324A4 (en) * | 2018-05-07 | 2020-08-20 | Apple Inc. | Multi-participant live communication user interface |
| CN110543857A (en) * | 2019-09-05 | 2019-12-06 | 安徽启新明智科技有限公司 | Contraband identification method, device and system based on image analysis and storage medium |
| CN113807119A (en) * | 2020-05-29 | 2021-12-17 | 魔门塔(苏州)科技有限公司 | Method and device for detecting person fixation position |
| CN112949461A (en) * | 2021-02-26 | 2021-06-11 | 北京高途云集教育科技有限公司 | Learning state analysis method and device and electronic equipment |
| CN114120432A (en) * | 2021-11-17 | 2022-03-01 | 湖北大学 | Online learning attention tracking method based on sight estimation and application thereof |
Non-Patent Citations (2)
| Title |
|---|
| BARZ, M: ""Automatic Visual Attention Detection for Mobile Eye Tracking Using Pre-Trained Computer Vision Models and Human Gaze"" * |
| 刘西: ""基于图片显著性的安检视觉搜索研究"" * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115620385B (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110348312A (en) | A kind of area video human action behavior real-time identification method | |
| JP2018113038A (en) | Inspection device, and method of detecting firearm in luggage | |
| US20080267458A1 (en) | Face image log creation | |
| CN107833221A (en) | A kind of water leakage monitoring method based on multi-channel feature fusion and machine learning | |
| CN101593425A (en) | A kind of fatigue driving monitoring method and system based on machine vision | |
| CN102085099B (en) | Method and device for detecting fatigue driving | |
| CN206193265U (en) | Cross-border electronic commerce intelligent security check system | |
| CN104077568A (en) | High-accuracy driver behavior recognition and monitoring method and system | |
| CN109685026A (en) | A kind of driver holds the method for real-time of mobile phone communication | |
| CN105913022A (en) | Handheld calling state determining method and handheld calling state determining system based on video analysis | |
| CN111563449A (en) | Real-time classroom attention detection method and system | |
| CN115841651B (en) | Constructor intelligent monitoring system based on computer vision and deep learning | |
| CN109063545A (en) | A kind of method for detecting fatigue driving and device | |
| CN114219687A (en) | Intelligent identification method for potential construction safety hazards by fusing human-computer vision | |
| CN112069907A (en) | X-ray machine image recognition method, device and system based on example segmentation | |
| CN115797856A (en) | Intelligent construction scene safety monitoring method based on machine vision | |
| CN115620385B (en) | Safety inspection staff attention detection method and system based on multivariate data | |
| CN115937793B (en) | Student behavior abnormality detection method based on image processing | |
| CN115375991A (en) | Strong/weak illumination and fog environment self-adaptive target detection method | |
| CN115331258A (en) | Industrial-grade applied danger forbidden zone personnel detection system and method | |
| CN107832698A (en) | Learning interest testing method and device based on array lens | |
| CN114898131A (en) | Contraband detection method and system based on natural guidance and data enhancement | |
| Ibrahim et al. | Mouth covered detection for yawn | |
| Wang et al. | Driving fatigue detection based on feature fusion of information entropy | |
| CN116894978B (en) | Online examination anti-cheating system integrating facial emotion and behavior multi-characteristics |
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: A method and system for attention detection of security personnel based on multivariate data Granted publication date: 20230728 Pledgee: Bank of Changsha Limited by Share Ltd. Wangcheng branch Pledgor: Hunan Ke Ke Intelligent Technology Co.,Ltd. Registration number: Y2024980001307 |