CN-121354226-B - Tamper-resistant and traceable handwritten data double-track recording system and recording equipment

CN121354226BCN 121354226 BCN121354226 BCN 121354226BCN-121354226-B

Abstract

The application relates to the technical field of image matching, in particular to a tamper-proof and traceable handwriting data double-track recording system and recording equipment, comprising an image acquisition module, a storage module and a data processing module, wherein the image acquisition module is used for acquiring a binarized electronic handwriting image, a retention image and an image to be verified of a user; the system comprises a paper handwriting extraction module, a key point evaluation module, a handwriting safety evaluation module and a handwriting safety evaluation module, wherein the paper handwriting extraction module is used for acquiring a binary paper handwriting image according to gray level change characteristics and texture complexity in an image to be verified, the key point evaluation module is used for recording the binary electronic handwriting image and the binary paper handwriting image as binary images, and screening out real key points in the binary images according to characteristics of continuous strokes in local windows of the key points in the binary images and the number of strokes connected with the key points, and the handwriting safety evaluation module is used for judging whether handwriting in the image to be verified is tampered. The application improves the accuracy of judging whether the handwriting is tampered or not by extracting the key points in the binary image more accurately.

Inventors

LIAO JUN
Zeng Yanyu
ZHOU LI
LIAO CHAO

Assignees

湖南诚翔信息技术有限公司

Dates

Publication Date: 20260508
Application Date: 20251107

Claims (7)

1. The tamper-resistant and traceable handwritten data double-track recording system is characterized by comprising the following modules: The image acquisition module is used for acquiring a binarized electronic handwriting image, a reserved image and an image to be verified of a user, wherein the reserved image is an image acquired by shooting a paper table after the user writes; The paper handwriting extraction module is used for extracting foreground points in the image to be verified, and acquiring handwriting confidence degrees of all the foreground points according to the difference degree between gradient directions of all the foreground points in a neighborhood window of each foreground point and the texture complexity degree of all the foreground points so as to acquire a binary paper handwriting image; The key point evaluation module is used for recording a binary electronic handwriting image and a binary paper handwriting image as binary images, acquiring all key points in each binary image, acquiring the confidence coefficient of a continuous stroke area of each key point according to the number of pixel points belonging to handwriting in a local window of each key point and the bending degree of edges in the local window of each key point, and acquiring the font characteristic index of each key point by combining the number of strokes connected with each key point, so as to screen out the real key points in each binary image; The handwriting safety evaluation module acquires a tampering index of the image to be verified according to a key point matching result of the image to be verified and the reserved image and a real key point matching result of the binarized electronic handwriting image and the binarized paper handwriting image, so as to judge whether handwriting in the image to be verified is tampered; The handwriting confidence coefficient acquisition method comprises the steps of acquiring gradient direction values of all foreground points in a neighborhood window of each foreground point, dividing the gradient direction values of all foreground points in the neighborhood window of each foreground point into two groups according to a preset gradient threshold value, and counting the average value of variances of the gradient direction values of the two groups; the method for acquiring the confidence coefficient of the continuous stroke area of each key point comprises the following steps: Counting the total number of handwriting pixel points in a local window of each key point, and recording the total number as the local density of each key point; Acquiring all edge pixel points in a local window of each key point, constructing a rectangular coordinate system by taking a first pixel point at the lower left corner of a binary image as an origin, marking the edge pixel point with the smallest Euclidean distance between the edge pixel points in the whole binary image as adjacent boundary points of each edge pixel point, calculating coordinate difference values between each edge pixel point and the adjacent boundary points to obtain displacement vectors, marking the included angle between each displacement vector and the positive direction of an x axis as the angle value of each edge pixel point, and counting the variance among the angle values of all edge pixel points in the local window of each key point; the confidence of the continuous stroke area of each key point respectively forms a positive correlation with the variance and forms a negative correlation with the local density; The method for acquiring the font characteristic index of each key point comprises the steps of generating a handwriting skeleton map with single pixel width of each binary image through a thinning algorithm, recording the handwriting skeleton map as a thinning image, mapping each key point in each binary image into a corresponding thinning image, recording the corresponding skeleton point as a mapping point of each key point if each key point is a skeleton point in the thinning image, recording the skeleton point closest to each key point as a mapping point of each key point in the corresponding thinning image if each key point is not the skeleton point, recording the total number of skeleton points in eight adjacent areas of the mapping points of each key point as the connection number of each key point, and recording the sum value of the normalization result of the confidence coefficient of the continuous stroke area of each key point and the normalization result of the connection number as the font characteristic index of each key point.
2. The tamper-resistant and traceable handwritten data dual-track recording system of claim 1, wherein foreground points in the image to be verified are pixels in the image to be verified with gray values greater than a preset foreground threshold.
3. The tamper-resistant and traceable handwriting data double-track recording system according to claim 1, wherein the binarization paper handwriting image obtaining method is characterized in that foreground points with handwriting confidence degrees larger than a preset handwriting threshold value in an image to be verified are marked as handwriting pixel points, gray values of all handwriting pixel points in the image to be verified are set to be 0, gray values of all other pixel points are set to be 255, and the obtained image is marked as the binarization paper handwriting image.
4. The tamper-resistant and traceable handwritten data dual-track recording system of claim 1, wherein the true keypoints in each binary image refer to keypoints in each binary image having a font characteristic index greater than or equal to a preset true threshold.
5. The tamper-resistant and traceable handwritten data double-track recording system according to claim 1 is characterized in that the tamper index of an image to be verified is obtained by generating key point descriptors of the image to be verified and a reserved image through an ORB algorithm, calculating the Hamming distance between the key point descriptors of the image to be verified and the reserved image, generating real key point descriptors of a binarized electronic handwriting image and a binarized paper handwriting image through the ORB algorithm, calculating the Hamming distance between the real key point descriptors of the two, and recording the average value of the obtained two Hamming distances as the tamper index of the image to be verified.
6. The system for dual track recording of handwritten data with tamper resistance and traceability as claimed in claim 1, wherein the specific process of judging whether the handwriting in the image to be verified is tampered is that if the tampering index of the image to be verified is smaller than a preset threshold, the handwriting in the image to be verified is judged not to be tampered, otherwise, the handwriting in the image to be verified is judged to be tampered.
7. A tamper-resistant and traceable handwritten data double-track recording device implementing the method as claimed in claim 1, characterized in that the handwritten data double-track recording device comprises a storage unit, a data acquisition unit, a data processing unit, a wireless communication unit, a wired communication interface, a mechanical component unit and a power supply; a storage unit for storing executable program codes and image data; The data acquisition unit comprises a pressure-sensitive handwriting board and a camera, and is used for acquiring an electronic handwriting image, a reserved image and an image to be verified of a user; a data processing unit for executing program code stored in the memory for analyzing whether handwriting in the image to be verified is tampered with; the wireless communication module is used for the double-track recording equipment to communicate with the server and the user mobile terminal; the wired communication interface is used for transmitting wired data; a mechanical part unit for adjusting the position or angle of the dual-track recording apparatus; a power supply for supplying power required for the dual-track recording apparatus.

Description

Tamper-resistant and traceable handwritten data double-track recording system and recording equipment Technical Field The application relates to the technical field of image matching, in particular to a tamper-proof and traceable handwritten data double-track recording system and a recording device. Background The electronic handwriting is a novel handwriting data protection means, however, the writing modes of the electronic handwriting and common paper handwriting are not consistent, so that the written handwriting has large difference, and mutual verification of the electronic handwriting and the paper handwriting cannot be realized. Therefore, the handwriting data recording system for ensuring that the handwriting is not tampered has important significance for the safety of the handwriting data. After the paper handwriting image and the electronic handwriting image are obtained, the two handwriting images are matched through an image matching technology, so that whether the handwriting is tampered or not is verified. The ORB algorithm is a classical image matching algorithm, has the characteristics of scale and rotation invariance, and can be well adapted to the problem of angular offset in paper image shooting. However, the ORB algorithm detects key points in the image by the FAST algorithm to achieve image matching. Because the electronic handwriting is a binary image generated by directly rendering the pressure sensitive data, and the FAST algorithm extracts the characteristic points through the gray value difference, the characteristic points extracted by the FAST algorithm in the binary image have larger errors, so that the accuracy of image matching is reduced, and the judging accuracy of whether the handwriting is tampered is further reduced. Disclosure of Invention In order to solve the technical problems, the application aims to provide a tamper-proof and traceable handwritten data double-track recording system and a recording device, and the adopted technical scheme is as follows: The embodiment of the application provides a tamper-resistant and traceable handwritten data double-track recording system, which comprises the following modules: the image acquisition module is used for acquiring a binarized electronic handwriting image, a retention image and an image to be verified of a user; The paper handwriting extraction module is used for extracting foreground points in the image to be verified, and acquiring handwriting confidence degrees of all the foreground points according to the difference degree between gradient directions of all the foreground points in a neighborhood window of each foreground point and the texture complexity degree of all the foreground points so as to acquire a binary paper handwriting image; The key point evaluation module is used for recording a binary electronic handwriting image and a binary paper handwriting image as binary images, acquiring all key points in each binary image, acquiring the confidence coefficient of a continuous stroke area of each key point according to the number of pixel points belonging to handwriting in a local window of each key point and the bending degree of edges in the local window of each key point, and acquiring the font characteristic index of each key point by combining the number of strokes connected with each key point, so as to screen out the real key points in each binary image; And the handwriting safety evaluation module is used for acquiring the tampering index of the image to be verified according to the key point matching result of the image to be verified and the retention image and the real key point matching result of the binarized electronic handwriting image and the binarized paper handwriting image, so as to judge whether the handwriting in the image to be verified is tampered. Preferably, the foreground point in the image to be verified refers to a pixel point in the image to be verified, where the gray value is greater than a preset foreground threshold value. The method for obtaining the handwriting confidence of each foreground point preferably comprises the steps of obtaining gradient direction values of all foreground points in a neighborhood window of each foreground point, dividing the gradient direction values of all foreground points in the neighborhood window of each foreground point into two groups according to a preset gradient threshold value, counting the average value of variances of the two groups of gradient direction values, obtaining texture values of all foreground points in the neighborhood window of each foreground point, counting the product of the variances and the average value of all texture values, and recording the sum value of the average value and the product as the handwriting confidence of each foreground point. The method for acquiring the binarized paper handwriting image comprises the steps of marking foreground points with handwriting confidence degrees larger than a preset handw