Worker registration machine and system based on recognition algorithm

Abstract

The invention discloses a worker registration machine and system based on an identification algorithm, and relates to the related field of information registration management, and the device comprises an interactive display module which is used for obtaining the basic information of workers; the face recognition module is used for obtaining identity information of a target worker; the certificate identification module is used for obtaining attribute information of a target worker; the registration verification module is used for obtaining a registration synchronization instruction; and the data synchronization module is used for synchronizing the target worker registration information to the worker management platform for information updating and sharing based on the registration synchronization instruction. The technical problems of long worker entry registration time and low management efficiency caused by many manual verification steps and low efficiency in the existing worker registration process are solved, and the technical effects of shortening the registration time and improving the management efficiency by realizing efficient and automatic identity verification and information synchronization through the self-service operation interface are achieved.

Description

Technical Field

[0001] The present application relates to fields related to information registration management, and in particular to a worker registration machine and system based on recognition algorithms. Background Art

[0002] In the current actual application of worker registration management, there are problems of cumbersome processes and low efficiency. Worker registration involves multiple steps of information entry, multiple verification of certificates and identities. Traditional registration methods usually require manual intervention and review, which is not only time-consuming, but also prone to errors or omissions in the process of information transmission and confirmation. In particular, with the increasing demand for intelligent site management, the demand for fast and accurate verification of worker identities and real-time information updates has become increasingly important. Since various types of registration information (such as basic information, facial information, certificate information, etc.) need to be collected, reviewed and synchronized in different systems, the existing worker registration process often requires cross-platform, multi-step operations, which not only increases management costs, but also may cause workers to wait too long and affect the construction progress. From worker self-registration to completion of entry, the traditional process conducts multiple reviews and matches of workers' basic information and identity information, which usually requires manual input and proofreading. In the process of data entry and comparison, it is easy to cause inaccurate or delayed information due to manual operation errors or incompatibility of information systems. Summary of the invention

[0003] This application provides a worker registration machine and system based on an identification algorithm to solve the technical problems in the existing worker registration process, such as long worker registration time and low management efficiency due to the large number of manual verification steps and low efficiency. It achieves the technical effect of realizing efficient and automated identity authentication and information synchronization through a self-service operation interface, thereby shortening the registration time and improving management efficiency.

[0004] The present application provides a worker registration machine based on a recognition algorithm, which includes: an interactive display module, which is used to display a worker self-service operation interface, and to interact with a target user through the worker self-service operation interface to enter information and obtain basic information about the worker; a face recognition module, which is used to adjust a face camera to collect a face image of the target user, obtain target face image information, and call a face recognition algorithm to perform face matching recognition on the face image information and a worker face database to obtain target worker identity information; a certificate recognition module, which is used to perform text recognition on a worker certificate set of the target user through an OCR device to obtain text information of the worker certificate, align and verify the text information of the worker certificate and the basic information of the worker, and obtain target worker attribute information; a registration verification module, which is used to obtain target worker registration information based on the target worker identity information and the target worker attribute information, and to review and verify the target worker registration information. When the information verification result passes, a registration synchronization instruction is obtained; and a data synchronization module, which is used to synchronize the target worker registration information to a worker management platform based on the registration synchronization instruction for information update and sharing.

[0005] The present application also provides a worker registration system based on a recognition algorithm, which is used for a worker registration machine based on a recognition algorithm. The system includes: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory.

[0006] This application proposes a worker registration machine and system based on recognition algorithm. Display the worker self-service operation interface, interact with the target user through the worker self-service operation interface to enter, and obtain the basic information of the worker; adjust the face camera to collect the face image of the target user, obtain the target face image information, call the face recognition algorithm to perform face matching recognition on the face image information and the worker face database, and obtain the target worker identity information; perform text recognition on the worker certificate set of the target user through the OCR device to obtain the text information of the worker certificate, align and verify the text information of the worker certificate and the basic information of the worker, and obtain the target worker attribute information; obtain the target worker registration information based on the target worker identity information and the target worker attribute information, review and verify the target worker registration information, and obtain the registration synchronization instruction when the information verification result passes; synchronize the target worker registration information to the worker management platform based on the registration synchronization instruction for information update and sharing. It solves the technical problems in the existing worker registration process, which is caused by many manual verification steps and low efficiency, resulting in long worker registration time and low management efficiency. It achieves the technical effect of realizing efficient and automated identity authentication and information synchronization through a self-service operation interface, thereby shortening the registration time and improving management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0008] Figure 1 A structural diagram of a worker registration machine based on a recognition algorithm provided in an embodiment of the present application.

[0009] Figure 2 A structural diagram of a worker registration system based on a recognition algorithm provided in an embodiment of the present application.

[0010] Explanation of the reference numerals: interactive display module 11 , face recognition module 12 , document recognition module 13 , registration and verification module 14 , data synchronization module 15 , processor 21 , memory 22 , input device 23 , output device 24 . DETAILED DESCRIPTION

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0012] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0013] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, device, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or components that are not clearly listed or inherent to these processes, devices, products, or equipment. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0014] The embodiment of the present application provides a worker registration machine based on a recognition algorithm, such as Figure 1 As shown, the device comprises:

[0015] The interactive display module 11 is used to display a worker self-service operation interface, and to interact with the target user through the worker self-service operation interface to obtain basic information of the worker.

[0016] Specifically, in the interactive display module 11, the system terminal displays the worker's self-service operation interface. The worker scans the QR code at the entrance or on the guidance page through the barcode scanner, and automatically identifies the QR code and preliminarily associates it with the worker's identity information, that is, synchronizes the worker's fixed information, including name, gender, date of birth, ID number, etc. After the scanning is completed, the preliminary registration page is displayed on the self-service operation interface, and the worker can enter or select relevant basic information through the interface, including place of residence, subcontractor information, contact information, etc. The interface also provides guidance prompts to ensure that the worker completes the information filling step by step. In this process, the recognition and verification function of the QR code greatly simplifies the entry steps of the worker's basic information, so that the worker does not need to manually enter the ID number or other complex information, which improves the convenience of operation and the accuracy of information entry. In this way, the worker can quickly complete the basic information entry of the self-service registration, laying the foundation for subsequent identity verification and registration review.

[0017] The face recognition module 12 is used to adjust the face camera to collect the face image of the target user, obtain the target face image information, call the face recognition algorithm to perform face matching and recognition on the face image information and the worker face database, and obtain the target worker identity information.

[0018] Specifically, in the face recognition module 12, the face camera will be adjusted according to the height and position of the target user to ensure that it is at the best shooting angle. Subsequently, the face camera is aimed at the worker's face for image acquisition to capture clear facial feature information. The collected facial image information will be transmitted back to the system terminal, and the system terminal will call the face recognition algorithm to extract features from the collected image information and compare it with the registered face data stored in the worker's face database. Through the matching analysis of the face recognition model, the system terminal confirms the identity information of the target worker in order to complete the subsequent registration process. Through the automated process, manual intervention is reduced, making identity authentication faster and more reliable.

[0019] Further, the adjusting the facial camera to collect facial images of the target user includes:

[0020] The panoramic image information of the target user is obtained through a panoramic camera, and the facial area recognition and selection are performed on the panoramic image information to obtain a target facial area frame diagram; based on the target facial area frame diagram, the facial height tracking and analysis of the target user is performed to obtain the target facial area height information; the facial recognition module 12 is communicatively connected with the lifting device, and the facial camera is mounted on the lifting device to obtain a camera height adjustment device; the facial camera is controlled by the camera height adjustment device according to the target facial area height information, and the camera height is adjusted and the facial image of the target user is captured.

[0021] Preferably, the system terminal uses a panoramic camera to take a full-body photo of the target user to obtain the user's panoramic image information. Based on this panoramic image information, the face area is identified and framed according to a face detection algorithm (such as a Haar feature cascade classifier, HOG+SVM or a deep learning model such as MTCNN, Faster R-CNN, etc.), and a rectangular frame diagram containing the face area is generated, i.e., a target face area frame diagram, for further processing. Subsequently, the face height of the target user is tracked and analyzed to determine the specific height position of the face relative to the ground. In this process, the system terminal obtains the installation height of the panoramic camera, and calculates the vertical position of the center point of the face area according to the vertical coordinate information of the target face area frame diagram, that is, the top coordinate and the bottom coordinate of the face area are averaged, and then the vertical coordinate of the vertical position of the center point of the face area is calculated with the total height pixel value of the panoramic image information of the body, and the relative height ratio of the center point of the face area in the image is obtained. After that, the difference is calculated by using 1 and the relative height ratio, and then multiplied by the installation height of the panoramic camera to obtain the absolute height of the facial area, that is, the height information of the target facial area. After obtaining the height information of the target facial area, the system terminal will obtain the camera height adjustment device, which is a lifting device for carrying a face camera. The lifting device has a communication connection with the face recognition module 12 and can receive the height information of the target facial area calculated by the face recognition module 12. Through the camera height adjustment device, the system terminal can automatically adjust the height of the face camera according to the acquired face height information, so that the face camera is aimed at the face area of ​​the target user. After completing the height adjustment, the face camera will collect facial images of the target user, capture clear facial features, and obtain the target facial image information. These collected target facial image information will be used in the subsequent face recognition process for comparison and identity verification. This automated process effectively improves the accuracy and efficiency of face collection and provides high-quality data support for subsequent identity comparison and verification.

[0022] Further, obtaining the target worker identity information includes:

[0023] The facial image information is standardized according to an image preprocessing step to obtain standard facial image information, wherein the image preprocessing step includes noise filtering, grayscale correction and face scaling; a face recognition algorithm is called to extract facial features and perform face recognition training on the worker face database to build a face recognition model; the face recognition model is verified and optimized to obtain a face recognition optimization model; and facial matching recognition is performed on the standard facial image information based on the face recognition optimization model to obtain the identity information of the target worker.

[0024] Preferably, after obtaining the facial image information, the system terminal performs standardization processing on the collected facial image information according to the image preprocessing step. Among them, the image preprocessing step includes performing noise filtering processing on the collected facial image to remove random noise in the image to improve the clarity of the image. Common methods include median filtering or Gaussian filtering. The facial image is converted to a grayscale image, and the brightness and contrast are corrected so that the image has a consistent grayscale distribution under different lighting conditions and enhances the detail performance of the image. The grayscale image is scaled so that all facial images meet the same size requirements (such as 128x128 pixels). This step ensures the uniformity of different images and facilitates subsequent feature extraction and matching. After the above-mentioned standardization processing, the system terminal obtains standardized, uniformly sized standard facial image information, which provides a basis for subsequent recognition. Subsequently, the face recognition algorithm is called to extract features from the facial images in the worker face database. Deep learning models such as convolutional neural networks (CNN) (such as DeepFace, FaceNet) are usually used to extract multi-dimensional key feature points from the image to form a unique facial feature vector for each worker. Then, these facial feature vectors are encoded by face region division, a sample data set is constructed, and the face recognition model is trained. After the training is completed, the face recognition model is verified using a sample data set that was not used for training to evaluate the recognition accuracy of the model. During the verification process, the accuracy of the face recognition model on these data is calculated and compared with the preset accuracy expectations. If it is less than the preset accuracy expectation, the model will be optimized by adjusting model parameters (such as learning rate, number of network layers) or using data enhancement technology to reduce the misrecognition rate. Otherwise, the current model configuration is saved to obtain an optimized high-precision face recognition model, that is, a face recognition optimization model. After that, the standard face image information obtained by preprocessing is input into the face recognition optimization model for facial feature matching and recognition, and the identity information of the target worker is output in combination with the worker face database for subsequent identity authentication and management.

[0025] Further, obtaining the target worker identity information includes:

[0026] A convolutional neural network is used to extract multi-dimensional feature points from each worker's face in the worker face database to obtain a worker face feature vector set; face area division and encoding is performed based on the worker face feature vector set to obtain a face coding area feature data set; data enhancement and identity identification are performed on the face coding area feature data set to obtain a worker face feature sample data set; a generative adversarial network is designed according to a face recognition algorithm, and alternating loss training is performed on the worker face feature sample data set based on the generative adversarial network to construct the face recognition model.

[0027] Preferably, the system terminal uses a convolutional neural network (CNN) to process each face image in the worker face database, and extracts multi-dimensional face feature points through deep convolution. CNN extracts low-level features (such as edges, textures) and high-level features (such as facial structure, key point positions) of the image layer by layer to form a multi-dimensional feature vector representing the features of each face image. Each worker face image is converted into a unique feature vector, thereby obtaining a worker face feature vector set. Subsequently, based on the worker face feature vector set, the face image is partitioned and encoded to improve the comprehensiveness and robustness of recognition. The system terminal divides the face into several regions (such as face, mouth, eyes, etc.) and encodes each region separately. These partitions can ensure that the details of each facial feature are processed independently, so that the recognition algorithm has a higher sensitivity to changes in local features (such as expressions, postures). Each partitioned region after encoding contains its feature position and feature value, and these regional feature data will be integrated into a human face facial encoding regional feature data set. The encoded regional feature dataset is then augmented with data, such as rotation and translation, to expand the diversity of the dataset and simulate the changes in faces under different environments and angles, so as to improve the generalization ability of the model. The identity information of each sample is marked on the augmented data to ensure that it can be effectively classified and identified according to the identity label in subsequent training, thereby obtaining a worker facial feature sample dataset that includes facial features under various conditions. Afterwards, a generative adversarial network (GAN) is designed based on the face recognition algorithm to further improve the recognition accuracy and adaptability of the model. The generative adversarial network consists of two parts: the generator and the discriminator. The generator is composed of a face recognition algorithm. By receiving random noise input, it generates sample data similar to the facial features of real workers, while the discriminator is used to distinguish between real samples and generated samples. The training of GAN is achieved through the alternating optimization of the generator and the discriminator, that is, the generator is constantly adjusted to generate realistic feature data, trying to deceive the discriminator so that the discriminator thinks that these generated samples are real, while the discriminator enhances its ability to distinguish between real samples and generated samples through adjustments. In each training cycle, the system terminal randomly extracts real samples from the worker facial feature sample data set, and uses the generator to generate a batch of fake samples, which are input into the discriminator to calculate the discriminant loss. The parameters of the discriminator are updated through back propagation so that it can more accurately distinguish between true and false. Then, the parameters of the discriminator are frozen, fake samples are generated again and input into the discriminator, and the parameters of the generator are updated to improve the authenticity of the samples generated by the generator, making it difficult for the discriminator to distinguish. This alternating training allows the generator to gradually generate more authentic worker facial feature samples, and the discriminator improves its discrimination ability in the process. When the generator and the discriminator reach a state of equilibrium, that is, when the discriminator can no longer accurately distinguish between the generated samples and the real samples, the system terminal builds a face recognition model for accurate worker identity verification.By training the generative adversarial network, the system terminal effectively improved the recognition accuracy and robustness of the model when feature data was limited, ensuring the stability and reliability of worker facial recognition in a variety of environments.

[0028] Further, obtaining a worker facial feature sample dataset includes:

[0029] Partition key points of the face facial coding area feature data set are identified to determine a face partition key point set; a face partition image expansion transformation parameter set is obtained according to the face partition key point set and the face image expansion target; data transformation and amplification are respectively performed on the face facial coding area feature set according to the face partition image expansion transformation parameter set to obtain an amplified face partition image set; data expansion processing is performed on the face facial coding area feature data set based on the amplified face partition image set to obtain the worker facial feature sample data set.

[0030] Preferably, the system terminal performs key point recognition on each partition (such as eyes, nose, mouth, etc.) in the facial coding area feature data set to extract important feature points in these areas. These key points include specific locations such as the corners of the eyes, wings of the nose, and corners of the mouth, thereby forming a facial partition key point set containing the coordinates of the key points of each area. This step ensures that the feature points of each facial area are accurately located, laying the foundation for subsequent image expansion. Subsequently, a transformation parameter set is generated based on the facial partition key point set and the expansion requirements of the facial image. The goal of facial image expansion is to increase data diversity, so these transformation parameters may include rotation angles, scaling ratios, translation displacements, etc. For example, the eye area is rotated 15 degrees clockwise, or the mouth area is scaled 1.2 times, etc., so as to generate a set of transformation parameters for each partition. After obtaining the facial partition image expansion transformation parameter set, the system terminal applies these transformation parameters to each area in the original face facial coding area feature set. By performing data transformation operations on the eyes, nose, mouth and other areas, multiple transformed image versions are generated and combined into an augmented facial partition image set containing rich facial partition features. Afterwards, based on the augmented facial partition image set, the system terminal performs data expansion processing on the original face facial coding area feature dataset, and merges the newly generated augmented facial partition image set into the face facial coding area feature dataset to form a worker facial feature sample dataset. This augmented dataset contains facial area images at various angles, sizes and positions, which is used to improve the generalization ability of the model so that it can still accurately identify the identity of workers in scenes with various postures and expressions.

[0031] The certificate recognition module 13 is used to perform text recognition on the worker certificate set of the target user through an OCR device to obtain text information of the worker certificate, align and verify the text information of the worker certificate and the basic information of the worker, and obtain the attribute information of the target worker.

[0032] Specifically, in the certificate recognition module 13, the system terminal scans the worker certificates in the worker certificate set of the target user through the OCR device to capture a clear certificate image, and then pre-processes the certificate image, including denoising, contrast enhancement, graying and binarization, etc., to ensure image clarity, thereby improving the accuracy of text recognition. After the pre-processing is completed, the text on the certificate image is recognized. During the recognition process, the system terminal cuts the entire text into separate characters or words, and then converts each character into a corresponding number or letter through a pattern matching algorithm. After that, by aligning and combining the converted results, the key information in the certificate, including the worker's name, certificate number, date of birth, company name, etc., is identified, thereby obtaining the text information of the worker's certificate. After obtaining the text information of the worker's certificate, the system terminal will align and verify the identified certificate information with the basic information entered by the user in the self-service interface, and ensure the consistency of the two by comparing the key data in the certificate (such as name, ID number or work number, etc.) and the corresponding fields in the basic information item by item. If the alignment verification is correct, the system terminal confirms the worker's identity information and attribute characteristics, including name, work number, age, position, department, etc., and records it as the target worker's attribute information for subsequent identity verification and management. If the information is inconsistent, it will be marked as abnormal, prompting manual review or re-entry to ensure the accuracy and consistency of the final information.

[0033] Further, obtaining a worker facial feature sample dataset includes:

[0034] The worker ID set of the target user is scanned and photographed by an OCR device to obtain worker ID image information; the worker ID image information is binarized and character segmented to obtain ID character image information; the ID character image information is subjected to multi-dimensional feature extraction and character image matching recognition by a pattern matching algorithm to obtain ID character recognition information; the ID character recognition information is subjected to text alignment and combination to obtain the worker ID text information.

[0035] Preferably, the system terminal uses an OCR device to scan and shoot the worker certificates in the worker certificate set of the target user to obtain high-definition worker certificate image information. This process includes shooting the ID card, work card or other worker identity certificate to ensure that the text on the certificate is clearly visible and convenient for subsequent recognition. Subsequently, the worker certificate image information is binarized. The purpose of binarization is to highlight the text part in the image. By setting a threshold, the pixel values ​​in the image are converted to black and white, the background is set to white, and the text is set to black to enhance the clarity of the text. Through binarization, the system terminal can effectively eliminate the interference of the background and ensure that the text boundaries are clear. After completing the binarization, the system terminal performs character segmentation on the binarized image. In this process, the image is scanned line by line, and the boundaries of each character are detected to segment a separate character image. For example, the gaps or boundaries between letters and numbers are detected, and each character is extracted independently according to these boundaries to form a collection of document character image information, including all character images that need to be recognized. After character segmentation, the system terminal uses a pattern matching algorithm (such as Tesseract OCR, PaddleOCR, etc.) to identify the image information of each ID character. The pattern matching algorithm analyzes the shape and features (such as edges, contours, strokes, etc.) of each character through multi-dimensional feature extraction, and compares these features with the OCR system character library. The character library stores the multi-dimensional features of standard characters, so the features in the ID character image can be matched with the characters in the library one by one, and the specific characters can be identified to form the ID character image information. After that, the system terminal performs text alignment and combination on the identified ID character information. This process combines the character recognition information according to the preset typesetting rules (such as the name in the first line, the ID number in the second line, etc.), and generates complete text information after alignment. For example, name: Zhang San, ID number: 123456789, etc. Through this alignment and combination, complete worker ID text information that conforms to the ID format can be generated, which is convenient for subsequent data comparison and identity authentication.

[0036] Further, obtaining the document character image information includes:

[0037] The image information of the worker ID card is converted into a grayscale image to obtain a grayscale image of the worker ID card. An enhanced grayscale threshold is set according to the distribution information of the grayscale image of the worker ID card. The grayscale image of the worker ID card is binarized and enhanced based on the enhanced grayscale threshold to obtain a binary image of the worker ID card. Contour feature detection and character image segmentation are performed on the binary image of the worker ID card to obtain character image information of the ID card.

[0038] Preferably, the system terminal preprocesses the collected worker ID image information and converts the original image into a grayscale image through a standard grayscale formula. The purpose of grayscale is to simplify the image data and convert each pixel in the image into a single grayscale value, thereby reducing the complexity of image processing while maintaining the main contours and features of the image. The grayscale image only contains light and dark information, not color information, and is therefore suitable for subsequent image enhancement and feature extraction. Subsequently, the pixel distribution information in the generated worker ID grayscale image is analyzed and statistically analyzed to generate a grayscale histogram of the image. The grayscale histogram shows the frequency of occurrence of each grayscale value (0-255) in the image. The horizontal axis of the histogram represents the grayscale value (from black to white), and the vertical axis represents the number of pixels for each grayscale value. This histogram reflects the overall distribution of brightness in the image. Then identify the main grayscale distribution areas of the background and foreground (text). Normally, the background of the ID image will be at a higher grayscale value (whiter), while the text area will be at a lower grayscale value (blacker). By analyzing these areas, the distribution range of the background and text can be obtained. After that, the threshold is calculated using the Otsu method or the adaptive threshold algorithm. The Otsu method determines the optimal threshold by finding the gray value in the grayscale histogram that minimizes the intra-class variance and maximizes the inter-class variance, thereby calculating the enhanced grayscale threshold that can effectively separate the foreground and background. The role of the enhanced grayscale threshold is used for subsequent binarization processing, which can help highlight the text part and form a higher contrast between the text area and the background area, thereby improving the accuracy of text recognition. Then, based on the set enhanced grayscale threshold, the system terminal performs binarization enhancement processing on the grayscale image to generate a binarized image. During the binarization process, each pixel in the grayscale image is converted to black (text area) or white (background area) according to the enhanced threshold. The binary image generated in this way can highlight the outline of the text area, making the text area easier to segment and identify. After the binary image is generated, the image is subjected to contour feature detection to identify and mark the character boundaries in the image. Contour feature detection usually uses contour extraction algorithms (such as OpenCV's findContours function) to find the outer border of the character. The system terminal will detect along the black pixels (text area) in the binary image, find the boundary information of the text area, and form the outline of each character. Finally, based on the results of the outline detection, the characters are segmented. In this process, the characters in the binary image will be segmented into independent small images one by one according to the detected outline, that is, the document character image information. Each segmented small image contains a character for subsequent OCR module recognition. Through the above processing, the system terminal obtains clear, segmented document character image information, which is convenient for subsequent accurate character recognition.

[0039] The registration verification module 14 is used to obtain the target worker registration information according to the target worker identity information and the target worker attribute information, and to review and verify the target worker registration information. When the information verification result passes, a registration synchronization instruction is obtained.

[0040] Specifically, in the registration verification module 14, the system terminal integrates the identity information of the target worker with the attribute information of the target worker to generate complete registration information of the target worker. This information contains all necessary data for verification and management to ensure that the identity and attribute information of the worker are accurate. After the information is integrated, all data in the registration information of the target worker is automatically reviewed and verified to ensure the accuracy and consistency of the information. The review and verification process includes consistency check, validity check, integrity check, etc. The consistency check compares the matching degree of repeated fields in the identity information and attribute information, such as certificate number, name, etc., to ensure that the data sources are consistent. The validity of the identity information and attribute information is determined through the validity check, such as the legitimacy of the certificate number, the logical rationality of the date of birth, etc. The integrity check confirms whether all required fields in the registration information are filled in completely to ensure that there are no omissions. If there is a verification failure, the information will be marked as pending review status, indicating that manual review is required. If all pass the verification, a registration synchronization instruction will be automatically generated. This instruction is used to synchronize the registration information of the target worker to the worker management platform to ensure that all information in the system is updated synchronously for subsequent identity authentication and management.

[0041] The data synchronization module 15 is used to synchronize the target worker registration information to the worker management platform for information update and sharing based on the registration synchronization instruction.

[0042] Specifically, in the data synchronization module 15, the system terminal packages the complete registration information including the worker's identity and attributes into a data packet according to the registration synchronization instruction, and sends the data packet to the worker management platform through the data transmission protocol. After receiving the data packet, the worker management platform parses the synchronization data and updates each worker information to the corresponding database to replace or supplement the existing information. In this way, the worker management platform can instantly obtain the latest registration information of the worker and realize real-time updating and sharing of information. This synchronization process ensures the consistency of worker data in each management system, enables relevant personnel on the platform to access accurate worker information, and provides data support for subsequent access management, authority allocation, and work arrangements.

[0043] In the above, refer to Figure 1 A worker registration machine based on a recognition algorithm according to an embodiment of the present invention is described in detail. Figure 2 A worker registration system based on a recognition algorithm according to an embodiment of the present invention is described.

[0044] A worker registration system based on a recognition algorithm according to an embodiment of the present invention comprises: a processor 21, a memory 22, an input device 23 and an output device 24; the number of the processor 21 in the system may be one or more. Figure 2 Taking a processor 21 as an example, the processor 21, memory 22, input device 23 and output device 24 in the system may be connected via a bus or other means. Figure 2 The example of connecting through bus is taken in the following.

[0045] The memory 22 is a computer-readable storage medium that can be used to store software programs, computer executable programs and components, such as program instructions / components corresponding to a worker registration machine based on a recognition algorithm in an embodiment of the present invention. The processor 21 executes various functional applications and data processing of a worker registration machine based on a recognition algorithm by running the software programs, instructions and components stored in the memory 22.

[0046] A worker registration system based on a recognition algorithm provided by an embodiment of the present invention is used for a worker registration machine based on a recognition algorithm, and has corresponding functional components and beneficial effects of a worker registration machine based on a recognition algorithm.

[0047] Although the present application makes various references to certain components in the system according to the embodiments of the present application, any number of different components may be used and run on a user terminal and / or server, and the various units and components included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0048] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A worker registration machine based on recognition algorithm, characterized in that: The worker registration machine includes: An interactive display module is used to display a worker self-service operation interface, and to interact with the target user through the worker self-service operation interface to obtain basic information of the worker; A face recognition module is used to adjust the face camera to collect the face image of the target user, obtain the target face image information, call the face recognition algorithm to perform face matching recognition on the face image information and the worker face database, and obtain the target worker identity information; The certificate recognition module is used to perform text recognition on the worker certificate set of the target user through an OCR device to obtain text information of the worker certificate, align and verify the text information of the worker certificate and the basic information of the worker, and obtain the attribute information of the target worker; A registration verification module, used to obtain the target worker registration information according to the target worker identity information and the target worker attribute information, and to review and verify the target worker registration information. When the information verification result passes, a registration synchronization instruction is obtained; A data synchronization module is used to synchronize the target worker registration information to the worker management platform for information update and sharing based on the registration synchronization instruction.

2. A worker registration machine based on recognition algorithm as claimed in claim 1, characterized in that: The step of adjusting the facial camera to collect facial images of the target user includes: Acquire panoramic image information of the target user through a panoramic camera, perform facial area recognition and frame selection on the panoramic image information, and obtain a target facial area frame map; Based on the target face area frame diagram, the target user's face height tracking analysis is performed to obtain target face area height information; The face recognition module is in communication connection with the lifting device, the face camera is mounted on the lifting device, and a camera height adjustment device is obtained; The camera height adjustment device controls the face camera according to the target face area height information to adjust the camera height and collect facial images of the target user.

3. A worker registration machine based on recognition algorithm as claimed in claim 1, characterized in that: The obtaining of the target worker identity information includes: Performing standardization processing on the face image information according to an image preprocessing step to obtain standard face image information, wherein the image preprocessing step includes noise filtering, grayscale correction and face scaling; Calling a face recognition algorithm to extract facial features and perform face recognition training on the worker face database to build a face recognition model; Verifying and optimizing the face recognition model to obtain an optimized face recognition model; Based on the face recognition optimization model, facial matching recognition is performed on the standard face image information to obtain the identity information of the target worker.

4. A worker registration machine based on recognition algorithm as claimed in claim 3, characterized in that: The constructing of the face recognition model comprises: Using a convolutional neural network to extract multi-dimensional feature points from each worker's face in the worker face database to obtain a worker face feature vector set; Based on the worker's facial feature vector set, face area division and encoding are performed to obtain a face coding area feature data set; Performing data enhancement and identity identification on the facial coding region feature dataset of the human face to obtain a sample dataset of facial features of workers; A generative adversarial network is designed according to a face recognition algorithm, and alternating loss training is performed on the worker facial feature sample data set based on the generative adversarial network to construct the face recognition model.

5. A worker registration machine based on recognition algorithm as claimed in claim 4, characterized in that: The step of obtaining a worker facial feature sample dataset comprises: Performing partition key point recognition on the facial encoding region feature data set to determine a facial partition key point set; Obtaining a facial subarea image expansion transformation parameter set according to the facial subarea key point set and the facial image expansion target; According to the facial subarea image expansion transformation parameter set, the facial encoding area feature set of the human face is respectively transformed and amplified to obtain an amplified facial subarea image set; The facial encoding region feature dataset of the human face is expanded based on the augmented facial partition image set to obtain the worker facial feature sample dataset.

6. A worker registration machine based on recognition algorithm as claimed in claim 1, characterized in that: The obtaining of text information of worker certificates includes: Scan and photograph the worker ID set of the target user through an OCR device to obtain image information of the worker ID; Binarizing and character segmenting the worker ID image information to obtain ID character image information; Using a pattern matching algorithm to perform multi-dimensional feature extraction and character image matching recognition on the document character image information to obtain document character recognition information; The character recognition information of the certificate is aligned and combined to obtain the text information of the worker certificate.

7. A worker registration machine based on recognition algorithm as claimed in claim 6, characterized in that: The obtaining of the document character image information includes: Convert the worker ID image information into a grayscale image to obtain a grayscale image of the worker ID, and set an enhanced grayscale threshold according to distribution information of the grayscale image of the worker ID; Performing binary enhancement processing on the grayscale image of the worker ID based on the enhanced grayscale threshold to obtain a binary image of the worker ID; Contour feature detection and character image segmentation are performed on the binary image of the worker's ID card to obtain the character image information of the ID card.

8. A worker registration system based on recognition algorithm, characterized in that: The system is used for a worker registration machine based on a recognition algorithm as described in any one of claims 1 to 7, comprising: A memory for storing executable instructions; A processor is used to execute the executable instructions stored in the memory.

Images