A construction machine fingerprint biometric man-machine binding attendance method and system

Abstract

The application discloses a kind of engineering machinery fingerprint biological recognition man-machine binding attendance methods and systems, comprising: by the block calculation and correction of pressing stability degree to fingerprint image, dynamically generate block adaptive filter weighting coefficient, then combine adaptive two-dimensional Gabor filter and enhance fingerprint image block by block, effectively solve the differentiation blur and ghost interference problem generated by uneven pressing and vibration sliding of fingerprint in strong vibration environment of engineering machinery, overcome the defects that traditional global fixed filter intensity is easy to cause clear area excessive filtering, fuzzy area enhancement is insufficient and callus, oil stain interference leads to low recognition accuracy, man-machine binding attendance is unreliable, significantly improve fingerprint recognition success rate and authorized reliability, meet the safety supervision demand of engineering machinery construction site.

Description

Technical Field

[0001] This invention belongs to the field of image data processing technology and relates to a fingerprint biometric human-machine binding attendance method and system for engineering machinery. Background Technology

[0002] With the increasing demands for intelligent construction machinery and stricter safety supervision at construction sites, fingerprint biometric technology is widely used in large, high-risk equipment such as excavators, cranes, and loaders to authenticate operators, bind them to machines, and track attendance, thus preventing safety hazards such as unlicensed operation, violations, and proxy attendance. Existing fingerprint recognition systems often employ adaptive two-dimensional Gabor filtering algorithms based on fuzzy directional field correction for fingerprint image enhancement, which can improve image blur and recognition success rate to some extent. However, in the actual operating environment of construction machinery, the cab is constantly subjected to high-frequency, high-vibration conditions such as engine vibration and road bumps. During fingerprint collection, the operator's finger may slip relative to the collection panel, resulting in motion blur and edge ghosting in the fingerprint image. Simultaneously, the pressure applied by the operator's finger is unevenly distributed; the central area of ​​the fingerprint is pressed firmly and less blurred, while the edge area is pressed loosely and more blurred, resulting in significant differences in blur characteristics within the same fingerprint image.

[0003] Existing Gabor filtering algorithms apply a globally uniform filtering enhancement intensity to the entire image, which cannot adapt to differences in local blur levels. Over-filtering in clear central areas can lead to artifacts and damage to the true ridge structure; while insufficient enhancement and failure to eliminate blur in areas with severe edge ghosting are problematic. Furthermore, calluses and oil on operators' fingers often interfere with the system's assessment of pressure stability, further increasing the fingerprint recognition false rejection rate. This makes it difficult to meet the reliability requirements for human-machine binding and attendance tracking in construction machinery, hindering the strict implementation of safety supervision at construction sites. Summary of the Invention

[0004] The purpose of this invention is to solve the problems in the existing fingerprint enhancement algorithm, which uses a globally uniform filtering strength and cannot adapt to the differential blurring caused by uneven pressing of fingerprints under strong vibration environment of engineering machinery, and is easily affected by calluses and oil stains, resulting in low fingerprint recognition accuracy and unreliable human-machine binding attendance. The invention provides a fingerprint biometric human-machine binding attendance method and system for engineering machinery.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] A fingerprint biometric attendance method for engineering machinery includes:

[0007] Collect fingerprint image data when construction machinery operators press the fingerprints, and preprocess the collected fingerprint image data to obtain preprocessed fingerprint grayscale images;

[0008] The acquired fingerprint grayscale image is divided into several non-overlapping fingerprint image blocks; based on the grayscale change characteristics within each fingerprint image block, the initial pressing stability of each fingerprint image block is obtained;

[0009] Based on the consistency of ridge direction within each fingerprint image block, the initial pressure stability of the fingerprint image block is corrected to obtain the true pressure stability of each fingerprint image block.

[0010] Based on the actual pressing stability of each fingerprint image block, the vibration ghosting interference caused by pressing is suppressed, and the filtering intensity weighting coefficient at each fingerprint image block is obtained.

[0011] An adaptive two-dimensional Gabor filter enhancement algorithm carrying corresponding filter strength weighting coefficients is used to enhance the fingerprint grayscale image block by block. The enhanced fingerprint image features are extracted and matched with a legitimate template in the cloud to complete the human-machine binding and attendance authorization for engineering machinery.

[0012] A further improvement of the present invention is that:

[0013] Furthermore, the process involves collecting fingerprint image data from the operator of the construction machinery when pressing the fingerprint, and preprocessing the collected fingerprint image data. Specifically, fingerprint image data is continuously collected at a frequency of 60 frames per second, and the frame with the largest effective pixel area is selected as the original fingerprint image. The original fingerprint image is then subjected to grayscale dimensionality reduction processing based on a grayscale conversion algorithm to obtain a single-channel fingerprint grayscale image.

[0014] Furthermore, the preliminary pressing stability of each fingerprint image block is obtained based on the grayscale change characteristics within each fingerprint image block, specifically as follows:

[0015] Based on the Sobel edge detection operator, obtain the first... The gradient magnitude corresponding to each pixel within a fingerprint image block And the gradient magnitude of all pixels in the fingerprint image block. Perform summation;

[0016] Total number of pixels based on fingerprint image blocks The arithmetic mean of the gradient magnitudes of the fingerprint image blocks is obtained.

[0017] A preset compensation constant is used to add the arithmetic mean to the compensation constant, thereby obtaining the gradient mean baseline value;

[0018] Taking the reciprocal of the gradient mean baseline value yields the inverse representation value of the grayscale contrast.

[0019] Adding a negative sign to the inverse representation value and using it as an exponent, with the natural constant as the index. Perform exponential operations on the base to obtain the first... Initial pressure stability of individual fingerprint image segments .

[0020] Furthermore, the acquisition of the first Initial pressure stability of individual fingerprint image segments Specifically:

[0021]

[0022] in, Indicates the first The initial pressure stability of each fingerprint image block. This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block The magnitude of the gradient at each pixel; To prevent the occurrence of a minimal constant compensation term with a value of zero in the denominator calculation; Indicates the first The arithmetic mean of the gradient magnitudes of all pixels within a fingerprint image block.

[0023] Furthermore, the initial pressure stability of each fingerprint image block is corrected based on the consistency of ridge direction within that block to obtain the true pressure stability of each block, specifically as follows:

[0024] Get the Within the first fingerprint image block The local gradient direction angle of each pixel This characterizes the local orientation of the fingerprint texture at that pixel.

[0025] Pixel-based local gradient direction angle The sum of the sine and cosine directions of all pixels within a fingerprint image block is obtained respectively.

[0026] Based on the sum of the sine and cosine directions of all pixels within a fingerprint image block, a direction synthesis value representing the degree of convergence of the overall texture direction of the block is obtained.

[0027] Divide the direction synthesis value by the total number of pixels N in the block to obtain the first... Consistency of ridge direction in individual fingerprint image blocks ;

[0028] Based on the Consistency of ridge direction in individual fingerprint image blocks With the preset first hyperparameter The corrected baseline value is obtained;

[0029] Based on the Initial pressure stability of individual fingerprint image segments The baseline value is corrected, and a penalty correction is applied based on the consistency of the ridge direction; the penalty correction value is then normalized to obtain the first... The stability of the actual pressure of each fingerprint image segment .

[0030] Furthermore, the acquisition of the first Consistency of ridge direction in individual fingerprint image blocks Specifically:

[0031]

[0032] in, Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block Local gradient direction angle at each pixel;

[0033] Furthermore, the acquisition of the first The stability of the actual pressure of each fingerprint image segment Specifically:

[0034]

[0035] in, Indicates the first The actual pressure stability of each fingerprint image block Indicates the first The initial pressure stability of each fingerprint image block; Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; This is the first hyperparameter, and its existence is to prevent... Extremely small values ​​can cause the evaluation link for the actual pressure stability to fail. The function is used to force the result to normalize.

[0036] Furthermore, based on the actual pressing stability of each fingerprint image block, the vibration ghosting interference caused by pressing is suppressed, and the filter intensity weighting coefficient at each fingerprint image block is obtained, specifically:

[0037] The stability of the actual pressure on the fingerprint image blocks Negation is performed, followed by natural exponential function calculation to obtain the exponentially decaying term. ;

[0038] Based on the exponential decay term With the preset minimum filter strength constant , obtain the Filter intensity weighting coefficients for each fingerprint image block ;

[0039] In order to effectively counteract and completely suppress ghosting interference caused by strong mechanical vibration, the fingerprint image block should be assigned a larger filtering intensity weighting coefficient to compensate for heavy filtering.

[0040] Furthermore, the acquisition of the first Filter intensity weighting coefficients for each fingerprint image block Specifically:

[0041]

[0042] in, Indicates the first The weighted coefficients of the filtering intensity at each fingerprint image block Indicates the first The degree of real-time pressure stability of each fingerprint image block; To ensure the minimum strength adjustment constant, even under conditions of extremely high stability in actual pressing, a basic level of smoothing filtering is still maintained to eliminate shot noise at the sensor's underlying surface.

[0043] Furthermore, the adaptive two-dimensional Gabor filter enhancement algorithm, based on weighted coefficients of corresponding filter strength, enhances the fingerprint grayscale image block by block, extracts the features of the enhanced fingerprint image, and matches them with a legitimate template in the cloud to complete the human-machine binding and attendance authorization for the engineering machinery. Specifically:

[0044] The fundamental frequency parameters of the two-dimensional Gabor filter are uniformly set to fixed values, and the direction parameters are aligned with the tangent direction of the true ridge of each fingerprint image block.

[0045] The globally fixed filter gain in the existing standard Gabor filtering algorithm is abolished. Instead, the base filter gain of each fingerprint image block is weighted by the filter intensity weighting coefficient corresponding to that fingerprint image block. Multiply the results to obtain the actual filtering gain for the fingerprint image block, and construct an adaptive two-dimensional Gabor spatial filtering function.

[0046] An adaptive two-dimensional Gabor spatial filtering function is used to perform two-dimensional spatial convolution enhancement on the fingerprint grayscale image block by block to obtain the fingerprint enhancement image;

[0047] The core feature coordinate vector is extracted from the fingerprint enhancement image and matched with the features of a legal fingerprint template pre-stored in the cloud. When the matching score is greater than the preset security control threshold, the construction machinery is unlocked, human-machine binding is completed, and attendance records are generated. When the matching score is less than the security control threshold, the construction machinery is locked and an unauthorized access warning is triggered.

[0048] Furthermore, the core feature coordinate vector includes the coordinate vectors of the branch points and endpoints of the fingerprint ridge;

[0049] Furthermore, the step of extracting the core feature coordinate vector from the fingerprint enhancement image and performing feature matching with the pre-stored legitimate fingerprint template in the cloud specifically involves:

[0050] Obtain the position coordinates, orientation, and type information of each feature point to form the fingerprint core feature coordinate vector;

[0051] Based on the cloud-based legitimate fingerprint templates in the security cloud database, retrieve the legitimate fingerprint templates of personnel who have pre-registered and hold special operation certificates for construction machinery;

[0052] The currently extracted core feature coordinate vector is compared with the cloud-based legitimate fingerprint template to perform feature point comparison and similarity calculation, thus obtaining a matching score;

[0053] The system compares the matching score with a preset security threshold and outputs a result indicating whether the match was successful or not.

[0054] A fingerprint biometric human-machine binding attendance system for engineering machinery includes:

[0055] The preprocessing module collects fingerprint image data when the operator of the construction machinery presses the fingerprint, and preprocesses the collected fingerprint image data to obtain a preprocessed fingerprint grayscale image.

[0056] The acquisition module divides the acquired fingerprint grayscale image into several non-overlapping fingerprint image blocks; based on the grayscale change characteristics within each fingerprint image block, it obtains the preliminary pressing stability of each fingerprint image block.

[0057] The correction module corrects the initial pressing stability of each fingerprint image block based on the consistency of the ridge direction within each fingerprint image block, thereby obtaining the true pressing stability of each fingerprint image block.

[0058] The suppression module, based on the actual pressing stability of each fingerprint image block, suppresses the vibration ghosting interference caused by pressing, and obtains the filter intensity weighting coefficient at each fingerprint image block;

[0059] The enhancement module enhances the fingerprint grayscale image block by block based on an adaptive two-dimensional Gabor filtering enhancement algorithm carrying corresponding filter strength weighting coefficients, extracts the features of the enhanced fingerprint image and matches them with a legitimate template in the cloud, and completes the human-machine binding and attendance authorization for engineering machinery.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] This invention effectively solves the problems of differential blurring and ghosting interference caused by uneven pressing and vibration sliding of fingerprints in the strong vibration environment of construction machinery by calculating and correcting the pressing stability of fingerprint images in blocks, dynamically generating block-adaptive filtering weighting coefficients, and then combining adaptive two-dimensional Gabor filtering to enhance the fingerprint image block by block. It overcomes the shortcomings of traditional global fixed filtering intensity, which easily leads to over-filtering of clear areas, insufficient enhancement of blurry areas, and interference from calluses and oil stains, resulting in low recognition accuracy and unreliable human-machine binding attendance. It significantly improves the fingerprint recognition success rate and authorization reliability, and meets the safety supervision needs of construction machinery construction sites. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 This is a flowchart illustrating a fingerprint biometric human-machine binding attendance method for engineering machinery according to the present invention.

[0064] Figure 2 This is a schematic diagram of the structure of a fingerprint biometric human-machine binding attendance system for engineering machinery according to the present invention. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0066] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0067] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0068] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper," "lower," "horizontal," or "inner" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0069] Furthermore, the use of the term "horizontal" does not imply that the component must be absolutely horizontal, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0070] In the description of the embodiments of the present invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention according to the specific circumstances.

[0071] The present invention will now be described in further detail with reference to the accompanying drawings:

[0072] See Figure 1 This invention discloses a fingerprint biometric human-machine binding attendance method for engineering machinery, comprising:

[0073] S101, collect fingerprint image data when the operator of the construction machinery presses the fingerprint, and preprocess the collected fingerprint image data to obtain the preprocessed fingerprint grayscale image;

[0074] A high-frequency optical fingerprint acquisition module, deeply embedded within the intelligent control panel of the construction machinery's cab, collects the operator's fingerprint image data in real time. The acquisition module continuously acquires fingerprint image data at a frequency of 60 frames per second. The system calculates the effective pixel area of ​​the fingerprint contact region in each frame in real time and selects the frame with the largest effective pixel area as the original fingerprint image to be processed. Based on a grayscale conversion algorithm, this original fingerprint image undergoes grayscale reduction processing to remove color information, resulting in a single-channel fingerprint grayscale image.

[0075] S102, the acquired fingerprint grayscale image is divided into several non-overlapping fingerprint image blocks; based on the grayscale change characteristics within each fingerprint image block, the preliminary pressing stability of each fingerprint image block is obtained.

[0076] The entire fingerprint grayscale image is evenly divided into several fingerprint image blocks of 16×16 pixels each, which do not overlap.

[0077] In actual engineering machinery vibration environments, when the operator presses firmly into the center of the fingerprint sensor, the fingertip skin adheres tightly to the surface of the fingerprint acquisition module, resulting in minimal relative micro-slippage between them. At this point, a high grayscale contrast is maintained between the fingerprint ridges and valleys, and the numerical changes in the gradient space are extremely dramatic. Conversely, the loosely pressed edge areas are prone to slippage under strong mechanical vibration, causing severe ghosting as the grayscale values ​​of the ridges and valleys intermingle and merge, resulting in smoother local grayscale changes. Therefore, the larger the average gradient amplitude of all pixels within each fingerprint image block, the higher the local grayscale contrast, the smaller the relative slippage caused by mechanical vibration in that fingerprint image block, and the greater its initial pressing stability. Based on the above logic, the initial pressing stability of each fingerprint image block, based on the grayscale change characteristics within each block, is specifically as follows:

[0078] Based on the Sobel edge detection operator, obtain the first... The gradient magnitude corresponding to each pixel within a fingerprint image block And the gradient magnitude of all pixels in the fingerprint image block. Perform summation;

[0079] Total number of pixels based on fingerprint image blocks The arithmetic mean of the gradient magnitudes of the fingerprint image blocks is obtained.

[0080] A preset compensation constant is used to add the arithmetic mean to the compensation constant, thereby obtaining the gradient mean baseline value;

[0081] Taking the reciprocal of the gradient mean baseline value yields the inverse representation value of the grayscale contrast.

[0082] Adding a negative sign to the inverse representation value and using it as an exponent, with the natural constant as the index. Perform exponential operations on the base to obtain the first... Initial pressure stability of individual fingerprint image segments .

[0083] The acquisition of the first Initial pressure stability of individual fingerprint image segments Specifically:

[0084]

[0085] in, Indicates the first The initial pressure stability of each fingerprint image block. This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block The magnitude of the gradient at each pixel; To prevent the occurrence of a tiny constant compensation term with a value of zero during denominator calculation, it is preset to 0.001; Indicates the first The arithmetic mean of the gradient magnitudes of all pixels within a fingerprint image block; The larger the value, the more likely it is to be the first. The higher the local grayscale contrast within a fingerprint image block, the more firmly the area is pressed and the less microscopic relative slippage caused by mechanical vibration; when As the value increases, the fractional term becomes smaller overall. Combined with the negative sign, the overall exponent shifts towards a direction where the absolute value decreases and approaches zero. It exhibits a monotonically increasing trend, thus making the first The greater the initial pressure stability of each fingerprint image block, the higher the initial pressure stability; that is, the larger the arithmetic mean of the gradient magnitude, the higher the grayscale contrast, and the greater the initial pressure stability. The larger the value.

[0086] S103, Based on the ridge direction consistency feature within each fingerprint image block, the initial pressing stability of the fingerprint image block is corrected to obtain the true pressing stability of each fingerprint image block;

[0087] While the initial pressure stability of fingerprint image blocks can effectively quantify and distinguish blurred and clear areas under normal conditions, operators of construction machinery often engage in heavy physical labor for extended periods. Their fingers are prone to developing thick, dried calluses or residual oil stains after machine repairs. In these special cases, fingerprint image blocks containing calluses or oil stains exhibit a high degree of similarity to clear fingerprint image blocks under normal conditions: both display extremely high grayscale contrast and drastic gradient changes in the local image space. Consequently, areas containing calluses or oil stains also acquire extremely large average gradient amplitudes. If the initial pressure stability is relied upon in this situation, abnormal areas with calluses or oil stains, and severely blurred due to loose pressure, are easily misjudged as tightly pressed, extremely stable areas requiring no strong filtering. This directly results in these callus or oil stain areas not receiving sufficient filtering enhancement compensation, failing to recover the true ridge features from the severe vibration ghosting, and ultimately failing to accurately solve the technical problem of attendance recognition failure.

[0088] To completely resolve this hidden danger, it's crucial to understand the difference between this special case and the normal situation: In a normal, firmly pressed, clear fingerprint area, the fingerprint ridges exhibit highly regular directional periodicity, with significant anisotropy in local texture distribution and highly consistent ridge direction. However, in special areas containing calluses or oil, the local grayscale jumps are completely chaotic and isotropic, lacking any unified and coherent ridge direction. Therefore, further analysis of the ridge direction consistency characteristics within each fingerprint image block is necessary to perform deep penalty optimization on the indicator. Specifically, the lower the degree of ridge direction consistency within each fingerprint image block, the greater the likelihood that the block belongs to a callus or oil interference area, and the lower the reliability of the initial pressing stability. Strict numerical penalty suppression is required to ensure that the calculated ridge direction consistency after optimization is achieved. The lower the actual pressure stability of each fingerprint image block, the more stable it is. Based on the above progressive logic, the actual pressure stability of each fingerprint image block is obtained as follows:

[0089] Get the Within the first fingerprint image block The local gradient direction angle of each pixel This characterizes the local orientation of the fingerprint texture at that pixel.

[0090] Pixel-based local gradient direction angle The sum of the sine and cosine directions of all pixels within a fingerprint image block is obtained respectively.

[0091] Based on the sum of the sine and cosine directions of all pixels within a fingerprint image block, a direction synthesis value representing the degree of convergence of the overall texture direction of the block is obtained.

[0092] Divide the direction synthesis value by the total number of pixels N in the block to obtain the first... Consistency of ridge direction in individual fingerprint image blocks ;

[0093] Based on the Consistency of ridge direction in individual fingerprint image blocks With the preset first hyperparameter The corrected baseline value is obtained;

[0094] Based on the Initial pressure stability of individual fingerprint image segments The baseline value is corrected, and a penalty correction is applied based on the consistency of the ridge direction; the penalty correction value is then normalized to obtain the first... The stability of the actual pressure of each fingerprint image segment .

[0095] The acquisition of the first Consistency of ridge direction in individual fingerprint image blocks Specifically:

[0096]

[0097] in, Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block The local gradient direction and angle at each pixel; the more uniform and regular the direction of the fingerprint ridges within a fingerprint image block, the closer the gradient directions of each pixel, and the larger the combined result of the sine and cosine sum. The larger the value, the more chaotic the texture direction within the block, the more dispersed the direction, and the smaller the composite result. The smaller the value;

[0098] The acquisition of the first The stability of the actual pressure of each fingerprint image segment Specifically:

[0099]

[0100] in, Indicates the first The actual pressure stability of each fingerprint image block Indicates the first The initial pressure stability of each fingerprint image block; Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; This is the first hyperparameter, and its existence is to prevent... Extremely small values ​​can cause the evaluation link for the actual pressure stability to fail; the default value is 0.15. The function is used to force the result to normalize.

[0101] S104, based on the actual pressing stability of each fingerprint image block, suppress the vibration ghosting interference caused by pressing, and obtain the filter intensity weighting coefficient at each fingerprint image block;

[0102] Based on the actual pressure stability of each fingerprint image block, the adaptive two-dimensional Gabor filtering enhancement algorithm is precisely and adaptively calculated for the filter intensity weighting coefficient at each fingerprint image block. The strictly followed logic is as follows: the greater the actual pressure stability of each fingerprint image block, the more definitively it indicates that the pressure in that local area is extremely firm, and the phenomenon of relative slippage due to vibration is negligible. To absolutely prevent the over-sharpening effect caused by uniform filtering from destroying the original clear ridge features and producing false artifacts, a smaller filter intensity weighting coefficient should be assigned to that block to achieve slight filtering. Conversely, the smaller the actual pressure stability of each fingerprint image block, the more definitively it indicates that the pressure in that area is severely loose, the relative slippage due to vibration is huge, and the image ghosting is extremely severe. To implement strong countermeasures and completely suppress the ghosting interference caused by strong mechanical vibration, a larger filter intensity weighting coefficient should be assigned to that block to achieve heavy filtering compensation. Based on the above logic, the filter intensity weighting coefficient at each fingerprint image block is specifically as follows:

[0103] The stability of the actual pressure on the fingerprint image blocks Negation is performed, followed by natural exponential function calculation to obtain the exponentially decaying term. ;

[0104] Based on the exponential decay term With the preset minimum filter strength constant , obtain the Filter intensity weighting coefficients for each fingerprint image block ;

[0105] In order to effectively counteract and completely suppress ghosting interference caused by strong mechanical vibration, the fingerprint image block should be assigned a larger filtering intensity weighting coefficient to compensate for heavy filtering.

[0106] The acquisition of the first Filter intensity weighting coefficients for each fingerprint image block Specifically:

[0107]

[0108] in, Indicates the first The weighted coefficients of the filtering intensity at each fingerprint image block Indicates the first The degree of real-time pressure stability of each fingerprint image block; To ensure a minimum strength adjustment constant, even under conditions of extremely high stability during actual pressing, a basic level of smoothing filtering is maintained to eliminate shot noise at the sensor's underlying surface; the default value is 0.4. The larger the value, the more likely it is to be the first. The more firmly the fingerprint image is pressed in sections, the stronger its resistance to vibration blurring; exponential decay term As it decreases, the result of the addition operation becomes smaller, thus making the final result of the addition operation smaller. The filter intensity weighting coefficient at each fingerprint image block The smaller the value, the less artifacts are caused by over-filtering; conversely, The smaller the value, the more severe the ghosting blur. The attenuation of the exponential term is reduced, and the corresponding weighted coefficient of the calculated filter intensity will be larger, so as to eliminate vibration ghosting.

[0109] S105 uses an adaptive two-dimensional Gabor filter enhancement algorithm with corresponding filter strength weighting coefficients to enhance the fingerprint grayscale image block by block, extracts the features of the enhanced fingerprint image and matches them with a legitimate template in the cloud, and completes the human-machine binding and attendance authorization for engineering machinery.

[0110] The basic frequency parameters of the two-dimensional Gabor filter are uniformly set to fixed values, with a preset empirical value of 0.125. The direction parameters are aligned with the tangent direction of the true ridge of each fingerprint image block.

[0111] The globally fixed filter gain in the existing standard Gabor filtering algorithm is abolished. Instead, the base filter gain of each fingerprint image block is weighted by the filter intensity weighting coefficient corresponding to that fingerprint image block. Multiply the results to obtain the actual filtering gain for the fingerprint image block, and construct an adaptive two-dimensional Gabor spatial filtering function.

[0112] An adaptive two-dimensional Gabor spatial filtering function is used to perform two-dimensional spatial convolution enhancement on the fingerprint grayscale image block by block to obtain the fingerprint enhancement image;

[0113] The core feature coordinate vector is extracted from the fingerprint enhancement image and matched with the features of a legal fingerprint template pre-stored in the cloud. When the matching score is greater than the preset security control threshold, the construction machinery is unlocked, human-machine binding is completed, and attendance records are generated. When the matching score is less than the security control threshold, the construction machinery is locked and an unauthorized access warning is triggered.

[0114] The core feature coordinate vector includes the coordinate vectors of the branch points and endpoints of the fingerprint ridge;

[0115] The process of extracting the core feature coordinate vector from the fingerprint enhancement image and performing feature matching with a pre-stored legitimate fingerprint template in the cloud is as follows:

[0116] Obtain the position coordinates, orientation, and type information of each feature point to form the fingerprint core feature coordinate vector;

[0117] Based on the cloud-based legitimate fingerprint templates in the security cloud database, retrieve the legitimate fingerprint templates of personnel who have pre-registered and hold special operation certificates for construction machinery;

[0118] The currently extracted core feature coordinate vector is compared with the cloud-based legitimate fingerprint template to perform feature point comparison and similarity calculation, thus obtaining a matching score;

[0119] The system compares the matching score with a preset security threshold and outputs a result indicating whether the match was successful or not.

[0120] When the matching score is greater than the preset safety control threshold, the construction machinery is unlocked, human-machine binding is completed, and attendance records are generated. When the matching score is less than the safety control threshold, the construction machinery is locked and an unauthorized driving warning is triggered. Specifically: if the final feature matching score is greater than the preset safety control threshold of 85 points, it is clearly determined that the current personnel are legally identified and possess operating qualifications. At this time, the security integrated management and control platform will issue digital instructions to automatically unlock the engine ignition control unit of the corresponding construction machinery to complete the strong human-machine safety binding, and simultaneously generate encrypted logs and upload them to the cloud to complete standardized attendance records. If the matching score is less than the safety control threshold, the system immediately enters a locked and suspended state, forcibly refusing to start the machinery, and driving the alarm to trigger an unauthorized driving warning record, thereby thoroughly and rigorously standardizing the attendance security and operational safety procedures at the construction site.

[0121] See Figure 2 This invention discloses a fingerprint biometric human-machine binding attendance system for engineering machinery, comprising:

[0122] The preprocessing module collects fingerprint image data when the operator of the construction machinery presses the fingerprint, and preprocesses the collected fingerprint image data to obtain a preprocessed fingerprint grayscale image.

[0123] The acquisition module divides the acquired fingerprint grayscale image into several non-overlapping fingerprint image blocks; based on the grayscale change characteristics within each fingerprint image block, it obtains the preliminary pressing stability of each fingerprint image block.

[0124] The correction module corrects the initial pressing stability of each fingerprint image block based on the consistency of the ridge direction within each fingerprint image block, thereby obtaining the true pressing stability of each fingerprint image block.

[0125] The suppression module, based on the actual pressing stability of each fingerprint image block, suppresses the vibration ghosting interference caused by pressing, and obtains the filter intensity weighting coefficient at each fingerprint image block;

[0126] The enhancement module enhances the fingerprint grayscale image block by block based on an adaptive two-dimensional Gabor filtering enhancement algorithm carrying corresponding filter strength weighting coefficients, extracts the features of the enhanced fingerprint image and matches them with a legitimate template in the cloud, and completes the human-machine binding and attendance authorization for engineering machinery.

[0127] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for fingerprint biometric human-machine binding attendance tracking for engineering machinery, characterized in that, include: The fingerprint image data of the construction machinery operator when pressing is collected, and the collected fingerprint image data is preprocessed to obtain the preprocessed fingerprint grayscale image; The acquired fingerprint grayscale image is divided into several non-overlapping fingerprint image blocks; Based on the grayscale variation characteristics within each fingerprint image block, the initial pressure stability of each fingerprint image block is obtained; Based on the consistency of ridge direction within each fingerprint image block, the initial pressure stability of the fingerprint image block is corrected to obtain the true pressure stability of each fingerprint image block. Based on the actual pressing stability of each fingerprint image block, the vibration ghosting interference caused by pressing is suppressed, and the filtering intensity weighting coefficient at each fingerprint image block is obtained. An adaptive two-dimensional Gabor filter enhancement algorithm carrying corresponding filter strength weighting coefficients is used to enhance the fingerprint grayscale image block by block. The enhanced fingerprint image features are extracted and matched with a legitimate template in the cloud to complete the human-machine binding and attendance authorization for engineering machinery.

2. The fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 1, characterized in that, The process involves collecting fingerprint image data from the operator of the construction machinery when pressing the fingerprint, and preprocessing the collected fingerprint image data. Specifically, fingerprint image data is continuously collected at a frequency of 60 frames per second, and the frame with the largest effective pixel area is selected as the original fingerprint image. The original fingerprint image is then subjected to grayscale dimensionality reduction processing based on a grayscale conversion algorithm to obtain a single-channel fingerprint grayscale image.

3. The fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 2, characterized in that, The preliminary pressure stability of each fingerprint image block is obtained based on the grayscale change characteristics within each fingerprint image block, specifically as follows: Based on the Sobel edge detection operator, obtain the first... The gradient magnitude corresponding to each pixel within a fingerprint image block And the gradient magnitude of all pixels in the fingerprint image block. Perform summation; Total number of pixels based on fingerprint image blocks The arithmetic mean of the gradient magnitudes of the fingerprint image blocks is obtained. A preset compensation constant is used to add the arithmetic mean to the compensation constant, thereby obtaining the gradient mean baseline value; Taking the reciprocal of the gradient mean baseline value yields the inverse representation value of the grayscale contrast. Adding a negative sign to the inverse representation value and using it as an exponent, with the natural constant as the index. Perform exponential operations on the base to obtain the first... Initial pressure stability of individual fingerprint image segments .

4. The fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 3, characterized in that, The acquisition of the first Initial pressure stability of individual fingerprint image segments Specifically: in, Indicates the first The initial pressure stability of each fingerprint image block. This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block The magnitude of the gradient at each pixel; To prevent the occurrence of a minimal constant compensation term with a value of zero in the denominator calculation; Indicates the first The arithmetic mean of the gradient magnitudes of all pixels within a fingerprint image block.

5. A fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 4, characterized in that, The initial pressure stability of each fingerprint image block is corrected based on the consistency of ridge direction within that block to obtain the true pressure stability of that block. Specifically: Get the Within the first fingerprint image block The local gradient direction angle of each pixel This characterizes the local orientation of the fingerprint texture at that pixel. Pixel-based local gradient direction angle The sum of the sine and cosine directions of all pixels within a fingerprint image block is obtained respectively. Based on the sum of the sine and cosine directions of all pixels within a fingerprint image block, a direction synthesis value representing the degree of convergence of the overall texture direction of the block is obtained. Divide the direction synthesis value by the total number of pixels N in the block to obtain the first... Consistency of ridge direction in individual fingerprint image blocks ; Based on the Consistency of ridge direction in individual fingerprint image blocks With the preset first hyperparameter The corrected baseline value is obtained; Based on the Initial pressure stability of individual fingerprint image segments The baseline value is corrected, and a penalty correction is applied based on the consistency of the ridge direction; the penalty correction value is then normalized to obtain the first... The stability of the actual pressure of each fingerprint image segment .

6. A fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 5, characterized in that, The acquisition of the first Consistency of ridge direction in individual fingerprint image blocks Specifically: in, Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; Indicates the first Within the first fingerprint image block Local gradient direction angle at each pixel; The acquisition of the first The stability of the actual pressure of each fingerprint image segment Specifically: in, Indicates the first The actual pressure stability of each fingerprint image block Indicates the first The initial pressure stability of each fingerprint image block; Indicates the first The degree of consistency of ridge direction in individual fingerprint image blocks This represents the total number of pixels within each fingerprint image block; This is the first hyperparameter, and its existence is to prevent... Extremely small values ​​can cause the evaluation link for the actual pressure stability to fail. The function is used to force the result to normalize.

7. A fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 6, characterized in that, The method for suppressing vibration ghosting interference caused by pressing, based on the actual pressing stability of each fingerprint image block, yields the filter intensity weighting coefficient for each fingerprint image block, specifically: The stability of the actual pressure on the fingerprint image blocks Negation is performed, followed by natural exponential function calculation to obtain the exponentially decaying term. ; Based on the exponential decay term With the preset minimum filter strength constant , obtain the Filter intensity weighting coefficients for each fingerprint image block ; In order to effectively counteract and completely suppress ghosting interference caused by strong mechanical vibration, the fingerprint image block should be assigned a larger filtering intensity weighting coefficient to compensate for heavy filtering. The acquisition of the first Filter intensity weighting coefficients for each fingerprint image block Specifically: in, Indicates the first The weighted coefficients of the filtering intensity at each fingerprint image block Indicates the first The degree of real-time pressure stability of each fingerprint image block; To ensure the minimum strength adjustment constant, even under conditions of extremely high stability in actual pressing, a basic level of smoothing filtering is still maintained to eliminate shot noise at the sensor's underlying surface.

8. A fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 7, characterized in that, The adaptive two-dimensional Gabor filter enhancement algorithm, based on weighted coefficients of corresponding filter strength, enhances the fingerprint grayscale image block by block, extracts the features of the enhanced fingerprint image, and matches them with a legitimate template in the cloud to complete the human-machine binding and attendance authorization for engineering machinery. Specifically: The fundamental frequency parameters of the two-dimensional Gabor filter are uniformly set to fixed values, and the direction parameters are aligned with the tangent direction of the true ridge of each fingerprint image block. The globally fixed filter gain in the existing standard Gabor filtering algorithm is abolished. Instead, the base filter gain of each fingerprint image block is weighted by the filter intensity weighting coefficient corresponding to that fingerprint image block. Multiply the results to obtain the actual filtering gain for the fingerprint image block, and construct an adaptive two-dimensional Gabor spatial filtering function. An adaptive two-dimensional Gabor spatial filtering function is used to perform two-dimensional spatial convolution enhancement on the fingerprint grayscale image block by block to obtain the fingerprint enhancement image; The core feature coordinate vector is extracted from the fingerprint enhancement image and matched with the features of a legal fingerprint template pre-stored in the cloud. When the matching score is greater than the preset security control threshold, the construction machinery is unlocked, human-machine binding is completed, and attendance records are generated. When the matching score is less than the security control threshold, the construction machinery is locked and an unauthorized access warning is triggered.

9. A fingerprint biometric human-machine binding attendance method for engineering machinery according to claim 8, characterized in that, The core feature coordinate vector includes the coordinate vectors of the branch points and endpoints of the fingerprint ridge; The process of extracting the core feature coordinate vector from the fingerprint enhancement image and performing feature matching with a pre-stored legitimate fingerprint template in the cloud is as follows: Obtain the position coordinates, orientation, and type information of each feature point to form the fingerprint core feature coordinate vector; Based on the cloud-based legitimate fingerprint templates in the security cloud database, retrieve the legitimate fingerprint templates of personnel who have pre-registered and hold special operation certificates for construction machinery; The currently extracted core feature coordinate vector is compared with the cloud-based legitimate fingerprint template to perform feature point comparison and similarity calculation, thus obtaining a matching score; The system compares the matching score with a preset security threshold and outputs a result indicating whether the match was successful or not.

10. A fingerprint biometric human-machine binding attendance system for engineering machinery, characterized in that, include: The preprocessing module collects fingerprint image data when the operator of the construction machinery presses the fingerprint, and preprocesses the collected fingerprint image data to obtain a preprocessed fingerprint grayscale image. The acquisition module divides the acquired fingerprint grayscale image into several non-overlapping fingerprint image blocks. Based on the grayscale variation characteristics within each fingerprint image block, the initial pressure stability of each fingerprint image block is obtained; The correction module corrects the initial pressing stability of each fingerprint image block based on the consistency of the ridge direction within each fingerprint image block, thereby obtaining the true pressing stability of each fingerprint image block. The suppression module, based on the actual pressing stability of each fingerprint image block, suppresses the vibration ghosting interference caused by pressing, and obtains the filter intensity weighting coefficient at each fingerprint image block; The enhancement module enhances the fingerprint grayscale image block by block based on an adaptive two-dimensional Gabor filtering enhancement algorithm carrying corresponding filter strength weighting coefficients, extracts the features of the enhanced fingerprint image and matches them with a legitimate template in the cloud, and completes the human-machine binding and attendance authorization for engineering machinery.

Images