A method and system for intelligent recognition of attendance behavior

Abstract

This invention proposes an intelligent recognition method and system for attendance behavior, belonging to the field of intelligent attendance technology. This method establishes a dynamic feature database to collect biometric information and location data of the target object. It then verifies the collected information using established attendance verification rules to determine whether the target object's attendance is satisfactory. This invention transforms the location information used in traditional attendance tracking into a series of location data collected by mobile phone attendance software as a reference, increasing its reference value and effectiveness, making location information a more reliable standard for attendance tracking. For face recognition, it divides the recognition into face recognition of the target object in the mobile phone attendance software and face recognition of the target object in the work area camera, ensuring the target object completes attendance normally. It proposes attendance rules that conform to this invention, according to which the system can quickly and accurately complete the attendance task for the target object.

Description

Technical Field

[0001] This invention relates to the field of intelligent attendance technology, and more specifically, to an intelligent recognition method and system for attendance behavior. Background Technology

[0002] With the digital transformation of enterprise management, attendance systems are gradually transitioning from traditional manual registration to intelligent technology. Currently, the mainstream intelligent attendance technologies are mainly divided into: biometric authentication technology, location-based check-in technology, and dynamic recognition technology based on the behavior of target objects in their work areas.

[0003] The existing intelligent attendance technologies make attendance convenient and fast, eliminating the need for staff to queue at designated locations every day to scan their faces or fingerprints for attendance; they can complete attendance using only a mobile app. However, a series of problems still exist, including: biometric recognition is easily affected by factors such as lighting, occlusion, and device resolution; location services are not accurate enough; and they have poor adaptability to abnormal scenarios. Therefore, a new intelligent attendance behavior recognition method needs to be developed to solve the above-mentioned problems. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned problems existing in the prior art and to greatly improve its technical effect on the basis of the original technology; this invention provides an intelligent recognition method for attendance behavior, the method comprising:

[0005] A dynamic feature library is established; the dynamic feature library is used to store personnel identification codes, biometric features, and location scatter information features; the personnel identification code is used to uniquely identify the user's identity, including: personnel identity information and employee number; the biometric features are used to identify the personnel's biometric features, including: facial feature vector and fingerprint feature vector; the location scatter information features are a travel scatter plot composed of a series of location scatter points, used to identify the travel features of personnel attendance; specifically, the steps for obtaining location scatter information features are as follows: S11, determine the attendance time, the attendance time refers to the time when work begins; and stipulate that the location scatter information of the target object is obtained within the two hours adjacent to the attendance time; S12, automatically start the positioning function of the mobile phone clock-in software within the two hours adjacent to the attendance time, and collect the location information of the target object every 2 minutes, the location information is the scatter information of the target object; S13, after the target attendance is qualified, record all the location information collected within the two hours adjacent to the attendance time to form the travel scatter plot of the target object, which is the location scatter information feature of the target object, and save the obtained location scatter information feature to the dynamic feature library.

[0006] The system uses a mobile phone attendance tracking app to obtain real-time biometric and location information of target individuals. The biometric information includes facial recognition, which is obtained through both the mobile phone attendance tracking app and the work area camera. The location information is obtained at regular intervals of 2 minutes using the mobile phone attendance tracking app's location function. Specifically, the biometric information refers to facial and fingerprint recognition. Facial recognition in the mobile phone attendance tracking app involves scanning the target individual's face using the app's recording function and recording facial features. Facial recognition in the work area camera involves acquiring image information of the target individual's work area using the work area camera and identifying the target's face in the image information using a target recognition algorithm. The method for obtaining the location information is as follows: the mobile phone attendance tracking app's location function is automatically activated within two hours prior to the attendance time, collecting the target individual's location information every 2 minutes. All location information collected within these two hours is combined to form the target individual's location information.

[0007] The obtained target object information is input into a dynamic feature database, and attendance verification rules are used to determine whether the target object has completed attendance. The attendance verification rules include: S31, comparing the obtained target object's location scatter information and biometric information with the corresponding dynamic features in the dynamic feature database; S32, if the location scatter information is within the corresponding location scatter information feature threshold and the biometric identification is valid, attendance is deemed satisfactory and no further verification is required; if either verification fails, attendance is deemed incomplete and further verification is required; the scatter information feature threshold is a location area generated based on the location scatter information of previously completed target objects; S33, if either the location scatter information or the biometric information verification fails, then... If the facial recognition of the target object in the work area camera is successful, the target object is deemed to have passed attendance; if it fails, the target object is deemed to have failed attendance. Specifically, the target object's information refers to biometric information and location scatter information. The biometric verification is considered valid if either facial recognition or fingerprint verification is performed in the mobile attendance software. The verification method is to compare the collected biometric information with facial and fingerprint feature vectors in a dynamic feature library, determining the similarity between the collected biometric information and the feature vectors in the library. If the similarity exceeds 85%, the biometric verification is considered valid. The failure of either verification refers to the verification of both the target object's biometric information and location scatter information. The steps for obtaining the feature threshold of scatter information are as follows: S320, continuously update the location scatter information features in the dynamic feature library using the location scatter information of the target object's attendance pass; S321, extract the location scatter information corresponding to each day from the location scatter information features, and record the same location scatter information for three days or more as reliable location scatter information based on the path formed by the location scatter information; calculate the average speed of the target object from the start of movement to attendance pass based on the time corresponding to each scatter point of the target object in the reliable location scatter information, where the average speed is the distance the target object moves per second; multiply the average speed of the target object by 30 seconds, and record it as the allowable translation distance of the location scatter information; the allowable translation distance of the location scatter information refers to the phase... For the scatter information feature threshold, the distance that the reliable location scatter information can be shifted is allowed; S322, connect the reliable location scatter information of the target object every day to form the target attendance path, and make a threshold reference path parallel to the target attendance path on each side of the target attendance path. The distance from each point on the threshold reference path to the target attendance path is equal to the allowed shift distance of the location scatter information. The position area between the two threshold reference paths is called the location scatter information feature threshold of the target object for the corresponding date; the threshold reference path refers to the path corresponding to the edge of the scatter information feature threshold of the target object for the corresponding date; S323, take the union of the location scatter information feature thresholds of all dates of the target object to generate the latest location scatter information feature threshold of the target object;During daily attendance checks, the location scatter plot information of the target object is compared with the latest location scatter plot information feature threshold.

[0008] In addition, the present invention also provides an intelligent recognition system for attendance behavior, which is used to implement the intelligent recognition method for attendance behavior described above. The system includes: an attendance information feature extraction module; an attendance information collection module; and an attendance information analysis module. The attendance information feature extraction module is used to extract and store the attendance information features of the target object. The extracted attendance information features include: personnel identification code, biometric features, and location scatter information features. The attendance information collection module is used to collect the attendance information of the target object. The collected attendance information of the target object includes: biometric information and location scatter information. The biometric information includes: face recognition information and fingerprint recognition information. The face recognition information includes: face recognition of the target object in the mobile phone check-in software and face recognition of the target object in the work area camera. The attendance information analysis module is used to analyze the information collected by the attendance information collection module according to specific rules, wherein the specific rules refer to the attendance verification rules from S31 to S33.

[0009] Furthermore, the aforementioned system also includes a computer central processing system; the computer central processing system includes a data storage center and a data analysis center; wherein, the data storage center is equivalent to an attendance information feature extraction module, which establishes a dynamic feature library of S10 by extracting and storing the attendance information features of the target object; the data analysis center is equivalent to an attendance information analysis module, which determines whether the attendance is qualified by comparing the collected information with the information in the dynamic feature library; finally, the attendance information collection module is the target object's mobile phone check-in software, which collects the target object's biometric information and location scatter information through the mobile phone check-in software, and transmits the collected information to the computer central processing system.

[0010] The beneficial effects of this invention are:

[0011] This invention provides an intelligent method and system for recognizing attendance behavior; it has the following advantages:

[0012] 1. This method transforms the location information from previous attendance check-in into a series of scattered location information collected by mobile check-in software as a reference; it increases the reference value and effectiveness, making location information a more reliable standard for attendance check-in.

[0013] 2. This method involves two aspects during face recognition: face recognition of the target object in the mobile phone attendance software and face recognition of the target object in the work area camera. It can prevent the target object from being recognized by the camera after the face recognition of the target object in the mobile phone attendance software fails, thus ensuring the normal attendance of the target object.

[0014] 3. This method proposes an attendance rule that conforms to the present invention. According to the rule, the system can quickly and accurately complete the attendance task of the target object. Attached Figure Description

[0015] Figure 1 This is a flowchart of an intelligent recognition method for attendance behavior according to the present invention.

[0016] Figure 2 This is a schematic diagram of an intelligent recognition system for attendance behavior according to the present invention. Detailed Implementation

[0017] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings; it should be understood that the specific embodiments given herein are only for illustration and explanation of the present invention and cannot be used to limit the present invention.

[0018] It should be noted that many specific details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention may have other embodiments and variations thereof. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0019] like Figure 1The flowchart shown is a method for intelligent recognition of attendance behavior according to an embodiment of the present invention. The flowchart includes: Step S10, establishing a dynamic feature library; the dynamic feature library is used to store personnel identification codes, biometric features, and location scatter information features; the personnel identification code is used to uniquely identify the user's identity, including: personnel identity information and employee number; the biometric features are used to identify the biometric features of personnel, including: facial feature vector and fingerprint feature vector; the location scatter information features are a travel scatter plot composed of a series of location scatter points, used to identify the travel features of personnel attendance; Step S20, obtaining the biometric information and location scatter information of the target object in real time through the personnel mobile phone check-in software; the facial recognition in the obtained biometric information is divided into: facial recognition of the target object in the mobile phone check-in software and facial recognition of the target object in the work area camera; the location scatter information is obtained at the same time intervals through the positioning function of the mobile phone check-in software. The interval is set to 2 minutes. Step S30 involves inputting the obtained target object information into a dynamic feature library and determining whether the target object has completed attendance according to attendance verification rules. These rules include: S31 comparing the obtained target object's location scatter information and biometric information with corresponding dynamic features in the dynamic feature library; S32 determining attendance is qualified and no further verification is needed if the location scatter information is within the corresponding location scatter information feature threshold and the biometric identification is valid; otherwise, attendance is deemed incomplete and requires further verification. The scatter information feature threshold is a location area generated based on the location scatter information of previously completed target objects; S33 activating face recognition of the target object in the work area camera if either location scatter information or biometric information fails verification. If recognition is successful, attendance is deemed qualified; otherwise, attendance is deemed unqualified.

[0020] In step S10, the steps for acquiring location scatter information features are as follows: S11, determine the attendance time, which refers to the time when work begins; and stipulate that the location scatter information of the target object will be acquired within the two hours preceding the attendance time; S12, automatically start the positioning function of the mobile phone clock-in software within the two hours preceding the attendance time, and collect the location information of the target object every 2 minutes, which is the scatter information of the target object; S13, after the target attendance is qualified, record all the location information collected within the two hours preceding the attendance time to form a scatter plot of the target object's journey, which is the location scatter information feature of the target object, and save the acquired location scatter information feature to the dynamic feature library.

[0021] In step S20, biometric information refers to the face recognition and fingerprint recognition of the target object; the face recognition of the target object in the mobile phone check-in software refers to scanning the face of the target object through the recording function of the mobile phone check-in software and recording the facial features of the face; the face recognition of the target object in the work area camera refers to obtaining the image information of the target object's work area through the camera of the work unit and recognizing the face of the target in the image information through the target recognition algorithm.

[0022] It should be noted that when performing facial recognition on a target object in the work area camera, the target object in the image information is first segmented using neural network technology; then, the face in the segmented target object is recognized.

[0023] In the above embodiment, specifically, the location scatter information is obtained by the location function of the mobile phone check-in software at the same time intervals: the location function of the mobile phone check-in software is automatically activated within two hours before the attendance time, and the location information of the target object is collected every 2 minutes. All the location information collected within two hours is combined to form the location scatter information of the target object.

[0024] In step S30, the obtained target object information is input into the dynamic feature library, which contains biometric information and location scatter information. Biometric validity means selecting either facial recognition or fingerprint recognition in the mobile check-in software for biometric identification. The identification method is to compare the collected biometric information with the facial feature vector and fingerprint feature vector in the dynamic feature library to determine the similarity between the collected biometric information and the feature vector in the library. If the similarity exceeds 85%, the biometric identification is considered valid. If either verification fails, it means that either the biometric information or the location scatter information fails verification. In other words, if either the biometric information or the location scatter information fails verification, it is determined that attendance has not been completed and further verification is required.

[0025] Further, the scatter information feature threshold includes the following steps: S320, continuously updating the location scatter information features in the dynamic feature library using the location scatter information of the target object's attendance pass; S321, extracting the location scatter information corresponding to each day from the location scatter information features, and recording the same location scatter information for three days or more as reliable location scatter information based on the path formed by the location scatter information; calculating the average speed from when the target object starts moving to when attendance is pass based on the time corresponding to each scatter point of the target object in the reliable location scatter information, where the average speed is the distance the target object moves per second; multiplying the average speed of the target object by 30 seconds, and recording it as... S322: The allowable translation distance of location scatter information; the allowable translation distance of location scatter information refers to the distance that the reliable location scatter information is allowed to be translated relative to the scatter information feature threshold; S322: Connect the reliable location scatter information of the target object every day to form the target attendance path, and make a threshold reference path parallel to the target attendance path on each side of the target attendance path. The distance from each point on the threshold reference path to the target attendance path is equal to the allowable translation distance of the location scatter information. The position area between the two threshold reference paths is called the location scatter information feature threshold of the target object for the corresponding date; the threshold reference path refers to the path corresponding to the edge of the scatter information feature threshold of the target object for the corresponding date;

[0026] S323, take the union of the location scatter information feature thresholds of the target object for all dates, and generate the latest location scatter information feature threshold of the target object; when checking attendance each day, compare the location scatter information of the target object with the latest location scatter information feature threshold.

[0027] like Figure 2 The diagram illustrates an intelligent attendance behavior recognition system according to the present invention. The system includes: an attendance information feature extraction module S100; an attendance information collection module S101; and an attendance information analysis module S102. The attendance information feature extraction module S100 extracts and stores attendance information features of the target object, including: personnel identification code, biometric features, and location scatter information features. The attendance information collection module S101 collects attendance information of the target object, including: biometric information and location scatter information. The biometric information includes: face recognition information and fingerprint recognition information. The face recognition information includes: face recognition of the target object in mobile phone check-in software and face recognition of the target object in a work area camera. The attendance information analysis module S102 analyzes the information collected by the attendance information collection module according to specific rules, where the specific rules refer to the attendance verification rules in S30.

[0028] In the above embodiments, specifically, the system further includes a computer central processing system; the computer central processing system includes a data storage center and a data analysis center; the data storage center is equivalent to the attendance information feature extraction module S100, which establishes a dynamic feature library of S10 by extracting and storing the attendance information features of the target object; the data analysis center is equivalent to the attendance information analysis module S102, which determines whether the attendance is qualified by comparing the collected information with the information in the dynamic feature library; finally, the attendance information collection module S101 is the mobile phone check-in software of the target object, which collects the biometric information and location scatter information of the target object through the mobile phone check-in software, and transmits the collected information to the computer central processing system.

Claims

1. An intelligent recognition method for attendance behavior, characterized in that, The method includes: S10, Establish a dynamic feature library; the dynamic feature library is used to store personnel identification codes, biometric features, and location scatter plot information features; the personnel identification code is used to uniquely identify the user's identity, including: personnel identity information and employee number; the biometric features are used to identify the personnel's biometric features, including: facial feature vector and fingerprint feature vector; the location scatter plot information features are a travel scatter plot composed of a series of location scatter points, used to identify the travel features of personnel attendance. S20, real-time acquisition of biometric information and location scatter information of target objects through personnel mobile phone check-in software; the facial recognition in the acquired biometric information is divided into: facial recognition of target objects in mobile phone check-in software and facial recognition of target objects in work area cameras; the location scatter information is acquired at the same time intervals through the positioning function of mobile phone check-in software, and the same time intervals are 2 minutes. S30: Input the obtained target object information into the dynamic feature library, and determine whether the target object has completed attendance according to the attendance verification rules; the attendance verification rules include: S31: Compare the obtained target object's location scatter information and biometric information with the corresponding dynamic features in the dynamic feature library; S32: If the location scatter information is within the corresponding location scatter information feature threshold and the biometric identification is valid, then attendance is deemed qualified and no further verification is required; if either verification fails, then attendance is deemed incomplete and further verification is required; the scatter information feature threshold is a location area generated based on the location scatter information of target objects that have previously completed attendance; S33: If either the location scatter information or the biometric identification information fails verification, then activate the target object's face recognition in the work area camera; if recognition passes, then the target object's attendance is deemed qualified; if it fails, then the target object's attendance is deemed unqualified. The steps for obtaining the scatter information feature threshold are as follows: S320, continuously update the location scatter information features in the dynamic feature library using the location scatter information of the target object's attendance pass; S321, extract the location scatter information corresponding to each day from the location scatter information features, and record the same location scatter information for three days or more as reliable location scatter information based on the path formed by the location scatter information; calculate the average speed of the target object from the start of movement to attendance pass based on the time corresponding to each scatter point of the target object in the reliable location scatter information, where the average speed is the distance the target object moves per second; multiply the average speed of the target object by 30 seconds, and record it as the allowable translation distance of the location scatter information; the allowable translation distance of the location scatter information refers to the allowable reliable location scatter information relative to the scatter information feature threshold. S322: Connect the reliable location scatter information of the target object each day to form the target attendance path. On each side of the target attendance path, draw a threshold reference path parallel to the target attendance path. The distance from each point on the threshold reference path to the target attendance path is equal to the allowable translation distance of the location scatter information. The position area between the two threshold reference paths is called the location scatter information feature threshold of the target object for the corresponding date. The threshold reference path refers to the path corresponding to the edge of the location scatter information feature threshold of the target object for the corresponding date. S323: Take the union of the location scatter information feature thresholds of all dates of the target object to generate the latest location scatter information feature threshold of the target object. When checking attendance each day, the location scatter information of the target object is compared with the latest location scatter information feature threshold.

2. The intelligent recognition method for attendance behavior according to claim 1, characterized in that, The dynamic feature library includes the following steps for acquiring location scatter information features: S11, determining the attendance time, which refers to the start time of work; and stipulating that the location scatter information of the target object be acquired within the two hours preceding the attendance time; S12, automatically activating the positioning function of the mobile phone clock-in software within the two hours preceding the attendance time, collecting the location information of the target object every 2 minutes, which is the scatter information of the target object; S13, after the target's attendance is qualified, recording all the location information collected within the two hours preceding the attendance time to form a scatter plot of the target object's journey, which is the location scatter information feature of the target object, and saving the acquired location scatter information feature to the dynamic feature library.

3. The intelligent recognition method for attendance behavior according to claim 1, characterized in that, The real-time acquisition of biometric information and location scatter information of the target object through the personnel mobile phone check-in software includes: biometric information refers to the target object's face recognition and fingerprint recognition; the target object's face recognition in the mobile phone check-in software refers to scanning the target object's face through the video recording function of the mobile phone check-in software and recording the facial features; the target object's face recognition in the work area camera refers to acquiring image information of the target object's work area through the work unit's camera and recognizing the target's face in the image information through a target recognition algorithm.

4. The intelligent recognition method for attendance behavior according to claim 1, characterized in that, The location scatter information is obtained at regular intervals through the positioning function of the mobile phone attendance software. This includes: automatically activating the positioning function of the mobile phone attendance software within two hours before the attendance time, collecting the location information of the target object every 2 minutes, and combining all the location information collected within two hours to form the location scatter information of the target object.

5. The intelligent recognition method for attendance behavior according to claim 1, characterized in that, The process of inputting the obtained target object information into the dynamic feature library includes: the target object information refers to biometric information and location scatter information; the biometric verification is valid, which includes: selecting either face recognition verification or fingerprint recognition verification of the target object in the mobile phone check-in software for biometric verification, and the verification method is: comparing with the face feature vector and fingerprint feature vector in the dynamic feature library, and determining the similarity between the collected biometric information and the feature vector in the feature library. If the similarity exceeds 85%, the biometric verification is considered valid; if any one of the verifications fails, it refers to the verification of the target object's biometric information and location scatter information.

6. An intelligent recognition system for attendance behavior, characterized in that, The system is used to implement the intelligent recognition method for attendance behavior as described in claim 1. The system includes: an attendance information feature extraction module; an attendance information collection module; and an attendance information analysis module. The attendance information feature extraction module is used to extract and store the attendance information features of the target object. The extracted attendance information features include: personnel identification code, biometric features, and location scatter information features. The attendance information collection module is used to collect the attendance information of the target object. The collected attendance information of the target object includes: biometric information and location scatter information. The biometric information includes: face recognition information and fingerprint recognition information. The face recognition information includes: face recognition of the target object in the mobile phone check-in software and face recognition of the target object in the work area camera. The attendance information analysis module is used to analyze the information collected by the attendance information collection module according to specific rules. The specific rules refer to the attendance verification rules of S30.

7. The intelligent recognition system for attendance behavior according to claim 6, characterized in that, The system also includes a computer central processing system; the computer central processing system includes a data storage center and a data analysis center; the data storage center is equivalent to an attendance information feature extraction module, which establishes a dynamic feature library of S10 by extracting and storing the attendance information features of the target object; the data analysis center is equivalent to an attendance information analysis module, which determines whether the attendance is qualified by comparing the collected information with the information in the dynamic feature library; finally, the attendance information collection module is the target object's mobile phone check-in software, which collects the target object's biometric information and location scatter information through the mobile phone check-in software, and transmits the collected information to the computer central processing system.

Images