CN-114943069-B - Method for realizing rapid management and authentication of real name system of assault worker through cloud network

CN114943069BCN 114943069 BCN114943069 BCN 114943069BCN-114943069-B

Abstract

The invention provides a cloud network-based rapid management and authentication method for an assault worker real-name system, which comprises the following steps of S1, enabling workers to enter an engineering site, obtaining worker types through the cloud network, triggering identity recognition operation for assault workers, S2, in the identity recognition operation process, collecting body contours of the assault workers through face recognition, adjusting contrast according to an external light environment, S3, conducting filtering operation on candidate frames of the assault workers, calculating an image operator, recognizing face and body contours according to assault worker relevance judgment results, and conducting double comparison by combining identity information collected by a safety helmet.

Inventors

HOU CHUNMING
Yan Qingyue
GONG HUAIDONG
Wu Yukan
DONG HUANHUAN

Assignees

中建三局四川建筑装备有限公司
中建三局集团有限公司

Dates

Publication Date: 20260508
Application Date: 20220517

Claims (6)

1. A cloud network realization assault worker real-name system rapid management authentication method is characterized by comprising the following steps: s1, a worker enters an engineering site, obtains the type of the worker through a cloud network, and triggers identity recognition operation aiming at an assault worker; S1-1, after assault workers arrive at a contracted construction project, the assault workers are brought into the management range of real-name personnel, personal identity cards are provided firstly, after information is read by an identity card reading device, identity information matching is carried out according to identity information prestored in a cloud network, the category of the workers is obtained, if the assault workers are assault workers, face information identification is carried out, the serial numbers of safety helmet chips are collected, and S1-2 of identity authentication of the assault workers is completed; S1-2, enabling assault workers with recorded identities to enter a project field, identifying people and caps through a gate system to be consistent, releasing and recording the entering time, otherwise, prohibiting the entering, enabling the assault workers to go to a designated area for operation after entering the field, enabling the whole operation process to be automatically identified by a nearby positioning base station at fixed time, recording time and place data, enabling the assault workers to go to the designated place for exiting after the operation is completed, identifying the people and caps through the gate system to be consistent, releasing and recording the exiting time, otherwise, prohibiting the exiting; s1-3, if the assault worker does not enter the field again for a plurality of continuous days, the system automatically judges that the worker is out of the field, if the worker does not enter the field for three continuous days, the system automatically judges that the worker is out of the field, and the out-of-field time is judged to be the last out-of-field record; s2, in the process of identity recognition operation, body contour collection is carried out on a plurality of assaults through face recognition, and contrast is adjusted according to the external light environment; s2-1, because of a plurality of engineering field personnel, different work types collect images and identities on the same authentication equipment, reduce noise for personnel in an image candidate frame, exclude invalid collected data, S2-2, acquiring body contour information of the assault worker to establish a body contour attribute value, Body contour attribute values Wherein For the height limit value in the candidate frame, the height limit value is used for eliminating the object which obviously does not accord with the height condition, Adjusting a feature factor for the candidate frame size; the number of candidate set samples formed for the image selected by the plurality of candidate frames in the image frame; S2-3, obtaining the attribute value of the facial feature Wherein B is a facial feature set, F x,y is the number of facial image region positions, A i is the differentiation parameter of the frontal facial image, C i is the differentiation parameter of the frontal facial image, For the facial feature recognition weights, Judging error adjustment factors for facial features; S3, filtering operation is carried out on the obtained candidate frames of the assaults, an image operator is calculated, face and body contour recognition is carried out according to the assaults relevance judging result, and double comparison is carried out by combining identity information collected by the safety helmet.
2. The cloud network implementation assault worker real-name system rapid management authentication method of claim 1, wherein the S2 further comprises: S2-4, adjusting the contrast ratio of the face and the background according to the initial value of the face feature, adjusting the contrast ratio of the background color and the face image, and adjusting the contrast ratio to be the controllable contrast ratio Z for recognizing the face after threshold value adjustment; U is the actual brightness value of the background color, V is the actual brightness value of the face, The brightness adjustment threshold is controllable, W is a preset brightness similarity threshold and is used for judging the similarity of background color and face brightness, so that difference adjustment is performed; judging the adjusting parameter according to the value of the controllable contrast Z The value of (2) is adopted, so that the contrast difference between the face and the background is regulated, and the face recognition rate is improved; ; A preset face contrast average value; Is a transition function, is used for performing function adjustment when judging that the contrast is different, and preventing overexposure or darkness, wherein T is a noise influence factor in the face image acquisition process.
3. The cloud network implementation assault worker real-name system rapid management authentication method of claim 1, wherein the S3 comprises: The step S3 comprises the following steps: S3-1, after obtaining the face image with the contrast adjusted, filtering a plurality of candidate frames of the face image and the body contour image, and improving the recognition degree of the face and the recognition degree of the body contour after filtering, wherein the filtering process is also a process of calculating a face filtering score and a body contour score, and the filtering score is obtained The definition is as follows: ; Wherein, the For the collection of facial features to be acquired, The face feature is obtained by the scale of The coordinates are The face candidate frame of (c) is reduced to a scale where only facial five sense organs are collected, The acquisition scale for body contour features is The coordinates are The body contour candidate frame of (1) is reduced to the dimension of only collecting four limbs, and the (4) is inner product operation.
4. The cloud network implementation assault worker real-name system rapid management authentication method of claim 3, wherein the S3 further comprises: s3-2, after calculating the filtering score, calculating pixel amplitude values of the face image and the body contour image, solving and obtaining a minimum operator of the sizes of candidate frames of the face image and the body contour image, ; Adjusting the value by the offset pixel according to the filtering score Combining the reordered x-axis pixel amplitude operation and the reordered y-axis pixel amplitude operation, thereby obtaining a facial image and a body contour image candidate frame which are adaptive to the minimum size.
5. The cloud network implementation assault worker real-name system rapid management authentication method of claim 4, wherein the step S3 further comprises: s3-3, the result of measuring the accurate value of the facial feature image is ; Wherein, by obtaining The maximum value of the number of the collected facial feature image change degrees is obtained, and then the facial images of assaults are collected in multiple time periods and angles, so that the most comprehensive facial images of the same assaults can be collected according to the following steps of Decision values are obtained for the face image, become noise reduction factors for judging the face image of the assault worker, And the control parameters are used for performing elimination operation on the obtained redundant candidate frames after the decision value of the face image result is obtained.
6. The cloud network implementation assault worker real-name system rapid management authentication method of claim 4, wherein the step S3 further comprises: the face characteristic information and the safety helmet authentication code information of the assault worker are obtained, and the face characteristic information and the safety helmet authentication code information are matched and consistent, so that the conditions of entering an engineering site can be met.

Description

Method for realizing rapid management and authentication of real name system of assault worker through cloud network Technical Field The invention relates to the field of identity authentication, in particular to a method for realizing rapid management and authentication of an assault worker real-name system by a cloud network. Background In China, the construction industry is one of the pillar industries of national economy, and develops along with the rapid development of economy. In recent years, the construction industry is increasingly obvious, the workers are forced to assault with the barren work, and the workers have the characteristics of short field time, uncertain operation time, high circulation rate, numerous people and the like, and cause huge management pressure on construction sites, so that the workers become the difficulty in management work of the workers. There is a need for a person skilled in the art to solve the corresponding technical problems. Disclosure of Invention The invention aims to at least solve the technical problems in the prior art, and particularly creatively provides a cloud network realization method for realizing rapid management and authentication of an assault worker real-name system. In order to achieve the above purpose of the present invention, the present invention provides a method for realizing fast management and authentication of a real name system of an assault worker by a cloud network, comprising the following steps: s1, a worker enters an engineering site, obtains the type of the worker through a cloud network, and triggers identity recognition operation aiming at an assault worker; s2, in the process of identity recognition operation, body contour collection is carried out on a plurality of assaults through face recognition, and contrast is adjusted according to the external light environment; S3, filtering operation is carried out on the obtained candidate frames of the assaults, an image operator is calculated, face and body contour recognition is carried out according to the assaults relevance judging result, and double comparison is carried out by combining identity information collected by the safety helmet. Preferably, the S1 includes: S1-1, after assault workers arrive at a contracted construction project, the assault workers are brought into the management range of real-name personnel, personal identity cards are provided firstly, after information is read by an identity card reading device, identity information matching is carried out according to identity information prestored in a cloud network, the category of the workers is obtained, if the assault workers are assault workers, face information identification is carried out, the serial numbers of safety helmet chips are collected, and S1-2 of identity authentication of the assault workers is completed; S1-2, enabling assault workers with recorded identities to enter a project field, identifying people and caps through a gate system to be consistent, releasing and recording the entering time, otherwise, prohibiting the entering, enabling the assault workers to go to a designated area for operation after entering the field, enabling the whole operation process to be automatically identified by a nearby positioning base station at fixed time, recording time and place data, enabling the assault workers to go to the designated place for exiting after the operation is completed, identifying the people and caps through the gate system to be consistent, releasing and recording the exiting time, otherwise, prohibiting the exiting; S1-3, if the assault worker does not enter the field again for a plurality of continuous days, the system automatically judges that the worker is out of the field, if the worker is not in the field for three continuous days, the system automatically judges that the worker is out of the field, and the out-of-field time is judged to be the last out-of-field record. Preferably, the S2 includes: S2-1, because of a plurality of engineering field personnel, different work types collect images and identities on the same authentication equipment, if the efficient batch collection is carried out, the personnel in the image candidate frame are required to be subjected to noise reduction, invalid collected data are eliminated, S2-2, acquiring body contour information of the assault worker to establish a body contour attribute value, Body contour attribute valuesWhereinFor the height limit value in the candidate frame, the height limit value is used for eliminating the object which obviously does not accord with the height condition,Adjusting a feature factor for the candidate frame size; the number of candidate set samples formed for the image selected by the number of candidate frames in the image frame. Preferably, the S2 further includes: S2-3, obtaining the attribute value of the facial feature Wherein B is a facial feature set, F x,y is the number of facial image region positions