Cloud network realizes the method for rapid management and authentication of sudden attack work real name system

Abstract

The application provides a cloud network implementation of a sudden worker real-name system fast management authentication method, including the following steps: S1, a worker enters an engineering site, obtains a worker type through a cloud network, and triggers an identity recognition operation for a sudden worker; S2, in the identity recognition operation process, a body contour of a plurality of sudden workers is collected through face recognition, and contrast is adjusted according to an external light environment; S3, a candidate frame of the obtained plurality of sudden workers is subjected to a filtering operation, an image operator is calculated, face and body contour recognition is performed according to a sudden worker correlation degree judgment result, and identity information collected by a safety helmet is subjected to double comparison.

Description

Technical Field

[0001] This invention relates to the field of identity authentication, and in particular to a cloud-based method for rapid real-name management and authentication of surprise workers. Background Technology

[0002] In my country, the construction industry, as one of the pillar industries of the national economy, has developed rapidly along with the economy. In recent years, the increasingly obvious labor shortage in the construction industry has led to the emergence of "rush workers." This type of worker is characterized by short on-site time, uncertain working hours, high turnover rate, and large numbers, which has created enormous management pressure on construction sites and become a difficult point in worker management. This necessitates that those skilled in the art solve the corresponding technical problems. Summary of the Invention

[0003] This invention aims to at least solve the technical problems existing in the prior art, and in particular, it innovatively proposes a cloud network-based method for rapid management and authentication of real-name registration for emergency workers.

[0004] To achieve the above-mentioned objectives of this invention, this invention provides a cloud-based method for rapid real-name management and authentication of assault workers, comprising the following steps:

[0005] S1, when a worker enters the construction site, the worker type is obtained through the cloud network, and an identity recognition operation is triggered for the worker who is arriving unexpectedly.

[0006] S2, during the identity recognition process, the body contours of several workers are collected through facial recognition, and the contrast is adjusted according to the external lighting environment.

[0007] S3 performs filtering operations on the candidate bounding boxes of several assault workers, calculates image operators, performs face and body contour recognition based on the correlation judgment results of the assault workers, and performs double comparison in combination with the identity information collected by the safety helmet.

[0008] Preferably, S1 includes:

[0009] S1-1: Once the emergency workers arrive at the agreed-upon construction project, they are immediately included in the real-name personnel management system. They first provide their personal ID cards, which are then read by the ID card reader. Based on the identity information pre-stored in the cloud network, the system matches the identity information to determine the worker's category. If the worker is an emergency worker, facial recognition and safety helmet chip number collection are performed to complete the emergency worker identity authentication and proceed to S1-2. If the worker is a regular worker, facial recognition and safety helmet chip number collection are not performed.

[0010] S1-2, registered emergency workers will enter the project site. They will be allowed entry and their entry time will be recorded if their hat and person match through the gate system. Otherwise, they will be prohibited from entering. After entering the site, the emergency workers will proceed to the designated area to work. During the entire work process, the chip information carried by the workers will be automatically identified by the nearby positioning base station at regular intervals, and the time and location data will be recorded. After the work is completed, the emergency workers will proceed to the designated location to leave the site. They will be allowed entry and their exit time will be recorded if their hat and person match through the gate system. Otherwise, they will be prohibited from leaving the site.

[0011] S1-3 If the worker does not return to the site for several consecutive days, the system will automatically determine that the worker has left the site. If the worker does not appear for three consecutive days, the system will automatically determine that the worker has left the site, and the time of departure will be considered as the last time the worker appeared.

[0012] Preferably, S2 includes:

[0013] S2-1. Due to the large number of personnel at the engineering site, and different job types collecting images and identities on the same authentication device, efficient batch collection requires noise reduction of personnel in the image candidate frames and elimination of invalid collected data.

[0014] S2-2, Obtain the body contour information of the assault workers and establish body contour attribute values.

[0015] Body contour attribute value ,in This is the height limit value in the candidate box. Objects that clearly do not meet the height requirement are excluded based on this height limit. The feature factor is used to adjust the candidate box size; The number of candidate set samples formed by selecting images from several candidate bounding boxes in an image frame.

[0016] Preferably, S2 further includes:

[0017] S2-3, Obtain facial feature attribute values Where B is the set of facial features, F x,y A represents the number of facial image region locations. i C represents the differential parameters for frontal facial images. i The parameters for distinguishing frontal facial images. Weights for facial feature recognition. This is a factor that adjusts for errors in facial feature judgment.

[0018] Preferably, S2 further includes:

[0019] S2-4, Adjust the contrast between the face and the background based on the initial value of the facial features. Since the face collection of the emergency workers is carried out outdoors or in places with unstable lighting, it is necessary to adjust the contrast between the background color and the face image. After threshold adjustment, it is adjusted to a controllable contrast Z for face recognition.

[0020] U represents the actual brightness value of the background color, and V represents the actual brightness value of the face. W is the controllable brightness adjustment threshold, and W is the preset brightness similarity threshold, which is used to judge the similarity between the background color and the brightness of the face, so as to make difference adjustment.

[0021] Determine the adjustment parameters based on the value of the controllable contrast ratio Z. The value of is adjusted to regulate the contrast difference between the face and the background, thereby improving the face recognition rate;

[0022] ;

[0023] The preset average facial contrast value; is the transition function used to adjust the function when there is a large difference in contrast, in order to prevent overexposure or underexposure. T is the noise factor in the face image acquisition process.

[0024] Preferably, S3 includes:

[0025] S3 includes:

[0026] S3-1: After obtaining the contrast-adjusted face image, filter the candidate bounding boxes of several face and body contour images. Filtering improves the recognition accuracy of the face and body contour. The filtering process also involves calculating the face filtering score and the body contour score. The definition is as follows:

[0027] ;

[0028] in, This is a collection of facial features. The scale for facial feature acquisition is The coordinates of the location are The face candidate bounding box was reduced to a scale that only captures facial features. To obtain the scale for body contour features The coordinates of the location are The body contour candidate box is reduced to a scale that only captures the limbs, and <,> are inner product operations.

[0029] Preferably, S3 further includes:

[0030] S3-2, after calculating the filtering score, pixel amplitude is calculated for the face image and body contour image, and the minimum candidate box size operator is obtained for the face image and body contour image.

[0031]

[0032] Based on the filtering score, the deviation pixel adjustment value By combining the magnitude calculations of the reordered x-axis pixels and the reordered y-axis pixels, candidate boxes for the face and body contour images with the smallest possible size are obtained.

[0033] Preferably, S3 further includes:

[0034] S3-3, the result of measuring the accuracy value of facial feature images is as follows:

[0035] ;

[0036] Among them, by obtaining To maximize the number of facial feature image variation samples collected, and thus collect facial images of assault workers from multiple time periods and angles, the most comprehensive facial images of the same assault worker can be obtained. The decision value for the facial image results becomes the noise reduction factor for judging the facial images of assault workers. It is the control parameter used to eliminate redundant candidate boxes after obtaining the decision value of the face image result.

[0037] Preferably, S3 further includes:

[0038] S3-3: Obtain the facial feature information and safety helmet authentication code information of the workers. Only when the facial feature information and safety helmet authentication code information match can they meet the conditions for entering the construction site.

[0039] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0040] It can securely and effectively authenticate the real names of emergency workers online, preventing confusion among personnel and improving authentication security. This is especially important in large-scale projects or construction sites, preventing unauthorized personnel from entering the construction site, ensuring project safety, and distinguishing emergency workers from regular workers. This enables efficient and precise management of emergency workers and provides strong support for protecting workers' rights.

[0041] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0042] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0043] Figure 1 This is a schematic diagram of the overall invention;

[0044] Figure 2 This is a flowchart of the workflow of the present invention. Detailed Implementation

[0045] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0046] like Figure 1 and 2 As shown, a method for rapid real-name management and authentication of assault workers using a cloud-based network is proposed, including the following steps:

[0047] S1, when a worker enters the construction site, the worker type is obtained through the cloud network, and an identity recognition operation is triggered for the surprise worker.

[0048] S2, during the identity recognition process, the body contours of several workers are collected through facial recognition, and the contrast is adjusted according to the external lighting environment.

[0049] S3 performs filtering operations on the candidate bounding boxes of several assault workers, calculates image operators, performs face and body contour recognition based on the correlation judgment results of the assault workers, and performs double comparison in combination with the identity information collected by the safety helmet.

[0050] refer to Figure 2 The logic flowchart of this invention patent includes information entry, identity identification, process monitoring, and seamless exit.

[0051] S1-1: Upon arrival at the designated construction project, the emergency workers are immediately included in the real-name personnel management system. They first provide their personal ID cards, which are then read by an ID card reader. The system matches the information with pre-stored identity information in the cloud network to determine the worker's category. If the worker is an emergency worker, facial recognition and helmet chip number collection are performed to complete the emergency worker identity authentication, proceeding to S1-2. If the worker is a regular worker, facial recognition and helmet chip number collection are not performed. A simple and convenient information entry method is used, collecting only ID card information for personnel registration. Simultaneously, the system links each worker to a unique chip inside their helmet, ensuring one helmet per person and clear identification. The system uses the chip in the helmet and positioning base stations deployed at the construction site to monitor and record the worker's work process, retaining accurate data for worker performance evaluation and wage payment.

[0052] S1-2, registered emergency workers will enter the project site. Entry will be permitted only if the gate system identifies the worker and their cap match, and the entry time will be recorded. Otherwise, entry will be prohibited. After entering, emergency workers will proceed to their designated work areas. Throughout the work process, nearby positioning base stations will periodically and automatically identify the chip information carried by the workers, recording the time and location data. After completing their work, emergency workers will proceed to their designated exit locations. Entry will be permitted only if the gate system identifies the worker and their cap match, and the exit time will be recorded. Otherwise, exit will be prohibited. Emergency workers will be identified within this system to distinguish them from ordinary workers. Once selected, they will enter the emergency worker real-name rapid management system.

[0053] S1-3: If a worker fails to return to the site for several consecutive days, the system automatically determines that the worker has left the site. If a worker fails to appear for three consecutive days, the system automatically determines that the worker has left the site, and the departure time is recorded as their last appearance. Correspondingly, the seamless exit process can be manually processed or automatically identified; if a worker fails to appear for three consecutive days, the system automatically determines that the worker has left the site, and the departure time is recorded as their last appearance.

[0054] S2 includes:

[0055] S2-1. Due to the large number of personnel at the engineering site, with different job types collecting images and identities on the same authentication device, efficient batch collection requires noise reduction of personnel within the image candidate frames to eliminate invalid data. During image acquisition, body contour determination is based on whether the candidate images within the candidate frames move and whether the width of the candidate frames changes during movement.

[0056] S2-2, Obtain the body contour information of the assault workers and establish body contour attribute values.

[0057] Body contour attribute value ,in This is the height limit value in the candidate box. Objects that clearly do not meet the height requirement are excluded based on this height limit. The feature factor for adjusting the candidate box size is used to acquire features of the selected image in the image frame with position coordinates x and y, and to judge the real-time changes of x and y. If there is a movement such as raising a hand, bending over, or turning to the side, the image in the candidate box will change, and the values ​​of x and y will also change in real time. The number of candidate set samples formed by selecting images from several candidate bounding boxes in an image frame;

[0058] S2-3, Obtain facial feature attribute values Where B is the set of facial features, F x,y A represents the number of facial image region locations. i C represents the differential parameters for frontal facial images. i The parameters for distinguishing frontal facial images. Weights for facial feature recognition. An adjustment factor for facial feature judgment errors; initial values ​​of facial features can be obtained by filtering features from the facial feature set;

[0059] S2-4, Adjust the contrast between the face and the background based on the initial value of the facial features. Since the face collection of the emergency workers is carried out outdoors or in places with unstable lighting, it is necessary to adjust the contrast between the background color and the face image. After threshold adjustment, it is adjusted to a controllable contrast Z for face recognition.

[0060] U represents the actual brightness value of the background color, and V represents the actual brightness value of the face. W is the controllable brightness adjustment threshold, and W is the preset brightness similarity threshold, which is used to judge the similarity between the background color and the brightness of the face, so as to make difference adjustment.

[0061] Determine the adjustment parameters based on the value of the controllable contrast ratio Z. The value of is adjusted to regulate the contrast difference between the face and the background, thereby improving the face recognition rate;

[0062] ;

[0063] The preset average facial contrast value; is the transition function, used to adjust the function when there is a large difference in contrast to prevent overexposure or underexposure; T is the noise influence factor in the face image acquisition process.

[0064] if When the value is zero, the acquired real-time facial contrast is less than the average facial contrast, and the transition function does not need to be adjusted, thus maintaining the current parameters for facial brightness and background color; if If the value is greater than zero, the real-time face contrast is greater than the average face contrast, so the face brightness and background color are adjusted; after adjustment, the contrast of the face and background colors is adapted.

[0065] S3 includes:

[0066] S3-1: After obtaining the contrast-adjusted face image, filter the candidate bounding boxes of several face and body contour images. Filtering improves the recognition accuracy of the face and body contour. The filtering process also involves calculating the face filtering score and the body contour score. The definition is as follows:

[0067] ;

[0068] in, This is a collection of facial features. The scale for facial feature acquisition is The coordinates of the location are The face candidate bounding box was reduced to a scale that only captures facial features. To obtain the scale for body contour features The coordinates of the location are The body contour candidate box is reduced to a scale that only captures the limbs, and <, > are inner product operations; for example, it is more difficult to capture a face in a 5×60 window, but it is easier to capture a face image in a 20×20 window.

[0069] S3-2, after calculating the filtering score, pixel amplitude is calculated for the face image and body contour image, and the minimum candidate box size operator is obtained for the face image and body contour image.

[0070] ;

[0071] Based on the filtering score, the deviation pixel adjustment value By combining the magnitude calculations of the reordered x-axis pixels and the magnitude calculations of the reordered y-axis pixels, candidate boxes for the face and body contour images with the smallest fit size are obtained.

[0072] S3-3, the result of measuring the accuracy value of facial feature images is as follows:

[0073] ;

[0074] Among them, by obtaining To maximize the number of facial feature image variation samples collected, and thus collect facial images of assault workers from multiple time periods and angles, the most comprehensive facial images of the same assault worker can be obtained. The decision value for the facial image results becomes the noise reduction factor for judging the facial images of assault workers. It is the control parameter for eliminating redundant candidate boxes after obtaining the decision value of the facial image result. Because the maximum number of facial feature image change samples is obtained through the continuous traversal and collection process of facial images, in all-weather engineering projects, the emergency workers frequently enter and exit the site. Although the most facial feature images are obtained, it is necessary to denoise invalid images. The decision value is determined based on the image result. This decision value is obtained through image training. After solving the measurement accuracy value of the facial feature image, in order to improve the accuracy of the emergency worker's identity, it is necessary to collect the safety helmet information of the emergency worker to obtain the pre-stored real name information of the emergency worker.

[0075] S3-3: Obtain the facial feature information and safety helmet authentication code information of the workers. Only when the facial feature information and safety helmet authentication code information match can they meet the conditions for entering the construction site.

[0076] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for rapid real-name management and authentication of emergency workers via cloud network, characterized in that, Includes the following steps: S1, when a worker enters the construction site, the worker type is obtained through the cloud network, and an identity recognition operation is triggered for the worker who is arriving unexpectedly. S1-1: Once the emergency workers arrive at the agreed-upon construction project, they are immediately included in the real-name personnel management system. They first provide their personal ID cards, which are then read by the ID card reader. Based on the identity information pre-stored in the cloud network, the system matches the identity information to determine the worker's category. If the worker is an emergency worker, facial recognition and safety helmet chip number collection are performed to complete the emergency worker identity authentication and proceed to S1-2. If the worker is a regular worker, facial recognition and safety helmet chip number collection are not performed. S1-2, The emergency workers whose identities have been registered will enter the project site. They will be allowed to enter if their hats match the gate system and the entry time will be recorded. Otherwise, they will be prohibited from entering. After entering the site, the emergency workers will go to the designated area to work. During the entire operation, the chip information carried by the workers will be automatically identified by the nearby positioning base station at regular intervals, and the time and location data will be recorded. After the work is completed, the workers will go to the designated location to leave the site. The gate system will recognize that the person and hat match and allow them to leave, and record the time of departure. Otherwise, they will be prohibited from leaving the site. S1-3, if the worker does not return to the site for several consecutive days, the system will automatically determine that the worker has left the site. If the worker does not appear for three consecutive days, the system will automatically determine that the worker has left the site, and the time of departure will be considered as the last time the worker appeared. S2, during the identity recognition process, the body contours of several workers are collected through facial recognition, and the contrast is adjusted according to the external lighting environment. S2-1: Due to the large number of personnel at the construction site, different job types are collecting images and identities on the same authentication device. Noise reduction is applied to personnel in the candidate image frames; invalid collected data is then eliminated. S2-2, Obtain the body contour information of the assault workers and establish body contour attribute values. Body contour attribute value ,in This is the height limit value in the candidate box. Objects that clearly do not meet the height requirement are excluded based on this height limit. The feature factor is used to adjust the candidate box size; The number of candidate set samples formed by selecting images from several candidate bounding boxes in an image frame; S2-3, Obtain facial feature attribute values Where B is the set of facial features, F x,y A represents the number of facial image region locations. i C represents the differential parameters for frontal facial images. i The parameters for distinguishing frontal facial images. Weights for facial feature recognition. Adjustment factor for facial feature judgment error; S3 performs filtering operations on the candidate bounding boxes of several assault workers, calculates image operators, performs face and body contour recognition based on the correlation judgment results of the assault workers, and performs double comparison in combination with the identity information collected by the safety helmet.

2. The cloud-based network-based method for rapid real-name management and authentication of assault workers according to claim 1, characterized in that, S2 further includes: S2-4, Adjust the contrast between the face and the background based on the initial value of the facial features, adjust the contrast between the background color and the face image, and adjust it to a controllable contrast Z for face recognition after threshold adjustment; U represents the actual brightness value of the background color, and V represents the actual brightness value of the face. W is the controllable brightness adjustment threshold, and W is the preset brightness similarity threshold, which is used to judge the similarity between the background color and the brightness of the face, so as to make difference adjustment. Determine the adjustment parameters based on the value of the controllable contrast ratio Z. The value of is adjusted to regulate the contrast difference between the face and the background, thereby improving the face recognition rate; ； The preset average facial contrast value; is the transition function used to adjust the function when there is a difference in contrast, to prevent overexposure or underexposure. T is the noise factor in the face image acquisition process.

3. The cloud-based network-based method for rapid real-name management and authentication of assault workers according to claim 1, characterized in that, S3 includes: S3 includes: S3-1: After obtaining the contrast-adjusted face image, filter the candidate bounding boxes of several face and body contour images. Filtering improves the recognition accuracy of the face and body contour. The filtering process also involves calculating the face filtering score and the body contour score. The definition is as follows: ； in, This is a collection of facial features. The scale for facial feature acquisition is The coordinates of the location are The face candidate bounding box was reduced to a scale that only captures facial features. To obtain the scale for body contour features The coordinates of the location are The body contour candidate box is reduced to a scale that only captures the limbs, and <,> are inner product operations.

4. The cloud-based network-based method for rapid real-name management and authentication of assault workers according to claim 3, characterized in that, S3 further includes: S3-2, after calculating the filtering score, pixel amplitude is calculated for the face image and body contour image, and the minimum candidate box size operator is obtained for the face image and body contour image. ； Based on the filtering score, the deviation pixel adjustment value By combining the magnitude calculations of the reordered x-axis pixels and the reordered y-axis pixels, candidate boxes for the face and body contour images with the smallest possible size are obtained.

5. The cloud-based network-based method for rapid real-name management and authentication of assault workers according to claim 4, characterized in that, S3 further includes: S3-3, the result of measuring the accuracy value of facial feature images is as follows: ； Among them, by obtaining To maximize the number of facial feature image variation samples collected, and thus collect facial images of assault workers from multiple time periods and angles, the most comprehensive facial images of the same assault worker can be obtained. The decision value for the facial image results becomes the noise reduction factor for judging the facial images of assault workers. It is the control parameter used to eliminate redundant candidate boxes after obtaining the decision value of the face image result.

6. The cloud-based network-based method for rapid real-name management and authentication of assault workers according to claim 4, characterized in that, S3 further includes: To qualify workers for entry into the construction site, their facial features and safety helmet authentication codes must be obtained and matched.

Images