Real estate information verification method and system based on image technology

Abstract

The invention discloses a real estate information verification method and system based on an image technology, and relates to the technical field of real estate information verification, and the method comprises the following steps: image collection: employing a collection device with multiple functions, and through the means of adjustable illumination, automatic paper flattening, multi-mode scanning, real-time quality monitoring, etc. Obtaining a high-quality and comprehensive paper house property certificate file image; acquiring a paper house property certificate file image by utilizing acquisition equipment which is provided with an adjustable lighting system and has an automatic paper flattening function; according to the method, comprehensive information is acquired through multi-mode scanning, the risk of information omission is effectively reduced, noise is accurately removed through image preprocessing, the contrast ratio is enhanced, information missing is repaired, a clear image is provided for follow-up processing, multiple kinds of accurate feature analysis are newly added in the feature extraction link, rich and accurate feature vectors are formed, and the image quality is improved. And the capability of judging the authenticity and integrity of the file is greatly enhanced, so that the accuracy and reliability of verification are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of real estate information verification, and specifically to a real estate information verification method and system based on image technology. Background Art

[0002] Real estate occupies an important position in people's lives and economic activities. The accuracy of real estate information is crucial for many aspects such as real estate transactions, mortgages, inheritances, etc. In real estate-related businesses, paper real estate certificates are the key basis for confirming property ownership and related information. Traditional real estate information verification mainly relies on manual methods. Staff need to manually check paper documents and compare database information. This method has many drawbacks. Manual verification is inefficient and time-consuming when dealing with a large number of real estate certificates. Moreover, manual operations are easily affected by subjective factors, and the judgment criteria of different verification personnel may vary, making it difficult to ensure the consistency of verification results. Long-term manual viewing is also prone to visual fatigue, which increases the probability of errors and affects the accuracy of verification. In addition, paper documents may be worn or damaged during storage and transmission, further increasing the difficulty of manual verification.

[0003] Paper real estate certificates themselves have certain vulnerabilities and risks of being tampered with. During long-term storage, the paper may be affected by natural environmental factors (such as humidity, light, etc.) and deteriorate, resulting in blurred handwriting, damaged paper, etc. This makes it difficult to accurately identify the information in the document, bringing great difficulties to the verification work. At the same time, with the development of technology, the means of forging and tampering with paper documents are becoming increasingly sophisticated. Some lawbreakers may produce false real estate certificates by forging seals, modifying text content, etc. Traditional manual verification methods are often unable to effectively identify these forging and tampering behaviors, easily allowing false documents to pass through, bringing serious security risks to the real estate trading market, damaging the legitimate rights and interests of all parties, and disrupting the market order.

[0004] To improve the verification efficiency and accuracy, image technology has gradually been applied to the field of real estate information verification. However, the existing verification methods based on image technology still have many deficiencies. In the image acquisition link, due to the diversity of paper documents (such as different paper quality, printing clarity, etc.) and possible wrinkles, stains, etc., the quality of the acquired images varies, affecting subsequent recognition and analysis. Moreover, traditional image acquisition devices and methods may not be able to fully obtain the hidden information in the document (such as anti-counterfeiting marks, watermarks, etc.), resulting in information omission. In the image analysis process, the method of comparing with the standard template is relatively single and not sensitive enough to some low-level errors generated by scanning and recognition (such as similar character confusion, partial information missing or repeated, etc.), making it difficult to detect and correct them in a timely manner. Summary of the Invention

[0005] The object of the present invention is to solve the problems of low efficiency, poor accuracy, easy damage and forgery of paper property certificate documents in manual verification, as well as poor quality of collected images, easy omission of low-level errors in analysis and lack of effective comparison of the same type of information in the existing image technology verification, and to propose a method and system for real estate information verification based on image technology.

[0006] The object of the present invention can be achieved by the following technical solutions:

[0007] A method for real estate information verification based on image technology, comprising the following steps

[0008] S1. Image acquisition: Using a collection device equipped with multiple functions, through adjustable lighting, automatic paper flattening, multi-mode scanning and real-time quality monitoring means, obtain high-quality and comprehensive images of paper property certificate documents;

[0009] S2. Image preprocessing: For the problems of noise, contrast and information loss in the collected images, use corresponding algorithms to perform noise reduction, contrast enhancement and information repair operations respectively to optimize the image quality;

[0010] S3. Feature extraction: On the basis of extracting conventional features, add logical structure analysis, refine text features and extract anti-counterfeiting element features to form rich and accurate file feature vectors;

[0011] S4. Information matching and verification: Compare the features of the document to be verified with those of the same type and standard template, statistically analyze the distribution law of similar information and use a machine learning model to identify error types, and at the same time dynamically calculate the similarity threshold for accurate judgment;

[0012] S5. Result output and report generation: Generate a verification report containing multi-faceted information, output it in multiple formats, mark and record suspicious information areas, send the report to manual review personnel and receive feedback for processing to complete the verification process.

[0013] Further, the specific steps of S1 are as follows:

[0014] Collect images of paper property certificate documents using a collection device equipped with an adjustable lighting system and an automatic paper flattening function; Detect the paper illumination situation through a sensor, and automatically adjust the illumination intensity and angle, and specifically adjust through the following formula: Where I adj is the illumination intensity adjustment value, k is the adjustment coefficient, set according to device characteristics and experience, I t is the target illumination intensity, I o is the illumination intensity of the paper surface currently detected, I maxIt is the maximum illumination intensity provided by the lighting device. According to the difference between the current illumination and the target illumination, the illumination intensity is adjusted proportionally based on this formula to ensure that the paper receives the correct illumination;

[0015] At the same time, a negative pressure adsorption or flexible flattening device is used to reduce paper wrinkles; three modes of optical scanning, infrared scanning, and ultraviolet scanning are adopted. Among them, optical scanning obtains the basic content of the document, infrared scanning detects anti-counterfeiting marks or watermarks hidden inside the paper, and ultraviolet scanning is for documents containing fluorescent anti-counterfeiting materials. After collection, the image data of the three modes are fused;

[0016] During the collection process, the image quality is monitored in real time through an image sharpness evaluation algorithm and a noise level detection method; among them, the image sharpness evaluation algorithm calculates the sharpness index C of the image through the variance method, where M and N are the number of rows and columns of the image respectively, and f(i, j) is the pixel value of the image at the coordinate (i, j), is the average pixel value of the image; the larger the sharpness index C, the clearer the image; a sharpness threshold C min is set. When C < C min , it is determined that the image sharpness is insufficient and needs to be recollected; if the image sharpness is insufficient, the image of this page of the document is automatically recollected;

[0017] For the salt-and-pepper noise caused by paper texture and stains in the collected image, the median filtering algorithm is used; for the Gaussian noise generated by uneven illumination or electronic device interference, the Gaussian filtering algorithm is adopted, and the filtering parameters are automatically adjusted according to the spatial distribution characteristics of the noise; A noise threshold N max is set. When N > N max , it is determined that there is too much noise and the image needs to be recollected.

[0018] Furthermore, the specific operation steps of S2 are as follows:

[0019] For the salt-and-pepper noise caused by paper texture and stains in the collected image, the median filtering algorithm is used; for the Gaussian noise generated by uneven illumination or electronic device interference, the Gaussian filtering algorithm is adopted, and the filtering parameters are automatically adjusted according to the spatial distribution characteristics of the noise;

[0020] The adaptive histogram equalization algorithm is used to enhance the image contrast to make the text and patterns clearer; for the information loss caused by paper damage or stains, it is repaired through an interpolation algorithm and a texture synthesis algorithm based on the neighborhood pixel information, and the corresponding algorithm parameters are selected according to the size of the missing area and the characteristics of the surrounding pixels.

[0021] Furthermore, the specific operation steps of S3 are as follows:

[0022] In addition to extracting the features of text, seals, and anti-counterfeiting markings, the analysis of the logical structure of the document content is added, including the relative positional relationship and arrangement order between different information blocks; the extraction of text features is refined, including the inclination angle of the text and the variation of stroke thickness.

[0023] For the microtext and rainbow printing pattern anti-counterfeiting elements in the paper real estate certificate document, their features are extracted using images and image analysis algorithms.

[0024] Analyze the size, spacing, arrangement pattern of the microtext and the color transition characteristics of the rainbow printing pattern; for the size and spacing characteristics of the microtext, they are quantified by calculating the pixel statistical information of the microtext area. Let the pixel value of the microtext area be m(i, j), where i = 1, 2,..., M w , j = 1, 2,..., N w , where M w and N w are the number of rows and columns of the microtext area respectively; the calculation of the average size S of the microtext is as follows: where n w is the number of valid pixels in the microtext area, which is determined by setting a pixel value threshold.

[0025] The microtext spacing D is obtained by calculating the average distance between adjacent microtext areas; first, the boundaries of the microtext areas are detected, then the distances between adjacent boundaries are calculated, and finally the average value is obtained to get the spacing D; these quantified features describe the microtext more precisely and are used for anti-counterfeiting verification.

[0026] Furthermore, the specific operation steps for calculating the text inclination angle in S3 are as follows:

[0027] The text inclination angle is calculated by determining the inclination angle of the minimum bounding rectangle of the text area. Specifically, let the pixel coordinates of the text area be (x i , y i ), where i = 1, 2,..., n and n is the number of pixel points in the text area; first, calculate the centroid coordinates (x c , y c ) of the text area: Then calculate the covariance matrix C:

[0028] Finally, by calculating the eigenvalues and eigenvectors of the covariance matrix, the inclination angle θ of the minimum bounding rectangle is obtained: where C 11 , C 12 , C 22 are the elements of the covariance matrix C; this angle is used as the text inclination feature for subsequent matching and analysis.

[0029] Further, the specific operation steps of S4 are as follows:

[0030] Compare the characteristics of the document to be verified with those of the same type of property certificate document and the standard template. By statistically analyzing the distribution rules of similar information in the same type of documents, including the property area range in the same region and the format of the property owner's name, preliminarily screen for suspicious information. Specifically:

[0031] Collect the information of existing and verified property certificate documents of the same type, including property area, property owner's name, property location, and information related to the building structure, and store them in the database according to the preset format and structure to ensure the accuracy and integrity of the data. At the same time, classify and sort the data according to the dimensions of region and property type.

[0032] Extract the characteristics of the document to be verified through the feature extraction step and form a complete feature vector of the document to be verified.

[0033] For the property area: Calculate the mean and standard deviation of the property areas of the same type, determine the reasonable range, and screen for suspicious information.

[0034] For the property owner's name: Screen relevant records from the database of property information of the same type, extract the property owner's name information; count all the surnames of the property owners, calculate the frequency of each surname; according to the statistical results, set the surname frequency threshold; extract the surnames of the property owners in the document to be verified and judge whether their frequencies are lower than the threshold; if lower than the threshold, judge that there is an abnormal situation, mark the document as suspicious in the format of the property owner's name, and record relevant information, including the document number, the property owner's name, and the frequency of this surname in the same type of documents.

[0035] Use a machine learning algorithm trained based on the data of verified documents to build a model to identify common error types in scanning and recognition, including similar character confusion, information repetition or missing error patterns, and assist in judging whether there are such errors.

[0036] When calculating the similarity between the document to be verified and the standard template and the property information of the same type, dynamically adjust the similarity thresholds of different characteristics according to factors such as document type, era, and region.

[0037] Further, the specific operation steps for adjusting the similarity thresholds of different characteristics in S4 are as follows:

[0038] After obtaining the file type, age, and regional factors, evaluate the file type, age, and regional factors respectively to obtain the type score, age score, and regional score. After normalization, use the sum of the type score and age score as the radius to establish the bottom circle, and use the regional score as the height to establish a cone model. Calculate the volume of this cone model and record it as the threshold value. Compare the obtained threshold value with several preset threshold value intervals. Different similarity thresholds are set corresponding to several preset threshold value intervals respectively. When determining the threshold value interval to which the threshold value belongs, determine the similarity threshold and substitute this similarity threshold into the calculation process of the similarity between the file to be verified and the standard template and the real estate information of the same type.

[0039] Further, the specific operation steps of S5 are as follows:

[0040] Generate a paper real estate certificate file information verification report according to the information matching and verification results. The report content includes the basic information of the file and the detailed situation of image acquisition, specifically the acquisition mode, quality assessment results, number of re-acquisitions, description of the preprocessing process, feature extraction results, comparison with the standard template and the real estate information of the same type, matching degree details, and the final verification conclusion; present the report, including displaying the change of feature similarity through charts and marking the suspicious information areas with images, and support multiple format outputs, including PDF, HTML, XML, for archiving and reference.

[0041] Further, the processing process of S5 for the suspicious information area is as follows:

[0042] Determine the annotation specifications for the suspicious information area, including the symbols, colors, shapes for clear annotation, and the format requirements for the annotation content; classify the emerging suspicious information, including seal types, text information types, and overall image types; for different types of suspicious information, formulate corresponding annotation methods and prompt information;

[0043] Use image processing software or annotation tools to draw rectangular frames on the image and add text annotation functions;

[0044] When generating the report, according to the suspicious information areas determined in the information matching and verification process, use the annotation tool to annotate according to the established specifications;

[0045] At the same time, establish a suspicious area information record form to record the location, suspicious type, relevant feature values, and corresponding file page numbers of each suspicious area; these recorded information will provide detailed data support for manual review together with the annotated images;

[0046] Send the generated report to the manual reviewers through the internal task assignment mechanism of the system or by email; when sending the email, mark it in the email subject and explain the importance of the report and the key points for review in the body of the email to remind the reviewers to pay special attention to the suspicious information areas; at the same time, provide a link to view the report or a way to download the attachment to ensure that the reviewers can obtain the report conveniently;

[0047] After receiving the report, the reviewers, in accordance with the specifications, view the content of the report and pay attention to the marked suspicious information areas; based on their professional knowledge and experience, conduct in-depth analysis of the suspicious areas; during the review process, use the review tools to record their review opinions and the problems found;

[0048] After the reviewers complete the review, they will feedback the review results; the review results include the final judgment on the suspicious areas, handling suggestions, and improvement suggestions for the verification system;

[0049] Update the verification status of the file according to the review results and perform corresponding subsequent operations according to the handling suggestions.

[0050] A real estate information verification system based on image technology, including:

[0051] An image acquisition module, which consists of a device control unit, a multi-mode scanning unit, and an image quality monitoring unit;

[0052] The device control unit controls the acquisition device, adjusts the lighting according to the lighting conditions, reduces paper wrinkles, and ensures the acquisition of high-quality images; the multi-mode scanning unit executes three scanning modes and fuses the data to obtain comprehensive file information; the image quality monitoring unit monitors the image quality in real time, judges according to the threshold and triggers re-acquisition to ensure that the images are qualified;

[0053] An image preprocessing module, which consists of a noise removal unit, an image enhancement unit, and an information repair unit;

[0054] The noise removal unit uses corresponding filtering algorithms and adjusts parameters for different noises to remove the noises; the image enhancement unit uses algorithms to enhance the image contrast for subsequent processing; the information repair unit repairs the information loss caused by damage or stains;

[0055] A feature extraction module, which consists of a conventional feature extraction unit, a logical structure analysis unit, a text refinement feature unit, and an anti-counterfeiting element feature unit;

[0056] The conventional feature extraction unit is used to extract the basic features of text and seals; the logical structure analysis unit is used to analyze the logical structure of the file content to assist in judgment; the text refinement feature unit is used to refine the text features to accurately depict the text; the anti-counterfeiting element feature unit extracts the anti-counterfeiting element features for anti-counterfeiting verification;

[0057] The information matching and verification module consists of a same-type data comparison unit, a machine learning model unit, and a similarity calculation unit;

[0058] The same-type data comparison unit is used to compare same-type data and screen for suspicious information such as property area and owner's name; the machine learning model unit uses the model to identify common types of scanning errors; the similarity calculation unit dynamically calculates the similarity threshold to accurately judge the file matching degree;

[0059] The result output and report generation module includes a report generation unit, a suspicious information annotation unit, and a report sending and feedback unit;

[0060] The report generation unit generates a verification report containing multi-faceted information and outputs it in multiple formats; the suspicious information annotation unit is used to mark the suspicious areas and record relevant information to assist manual review; the report sending and feedback unit sends the report to the manual review personnel, receives feedback and processes it to improve the verification process.

[0061] Compared with the prior art, the beneficial effects of the present invention are:

[0062] (1) In the present invention, multi-mode scanning is adopted to obtain comprehensive information, effectively reducing the risk of information omission. Image preprocessing accurately removes noise, enhances contrast, and repairs missing information, providing a clear image for subsequent processing. In the feature extraction link, a variety of accurate feature analyses are added, forming rich and accurate feature vectors, greatly enhancing the ability to judge the authenticity and integrity of the file, thus significantly improving the accuracy and reliability of verification;

[0063] (2) In the present invention, the optimization measures during image acquisition ensure the rapid acquisition of high-quality images. The real-time monitoring and automatic re-acquisition mechanism avoid ineffective work. The machine learning model helps to quickly identify common errors, shortening the error screening time. Dynamically adjusting the similarity threshold improves the accuracy and efficiency of matching judgment, and at the same time enhances the adaptability of the system to different types of files, effectively improving the overall verification efficiency;

[0064] (3) In the present invention, the generated detailed verification report is intuitively presented in multiple ways and supports multi-format output, optimizing information transmission and recording. The perfect annotation of the suspicious information area standardizes the processing process, facilitating the manual review personnel to quickly locate and understand the problem, improving the review efficiency. The close cooperation between the system and manual review realizes the complementary advantages, improves the real estate information verification work, effectively guarantees the safety of real estate transactions, and comprehensively optimizes the verification process. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings;

[0066] Figure 1 It is the system general block diagram of the present invention. Detailed implementation manners

[0067] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0068] It should be understood that the terms "including" and "comprising" used in the specification and claims of this disclosure indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0069] It should also be understood that the terms used in this disclosure specification are only for the purpose of describing specific embodiments and are not intended to limit this disclosure. As used in this disclosure specification and claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in this disclosure specification and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0070] As Figure 1 shown, a method for verifying real estate information based on image technology includes the following steps:

[0071] Image acquisition: Using a collection device equipped with multiple functions, through means such as adjustable lighting, automatic paper flattening, multi-mode scanning, and real-time quality monitoring, to obtain high-quality and comprehensive images of paper real estate certificate documents;

[0072] Collecting images of paper real estate certificate documents using a collection device equipped with an adjustable lighting system and an automatic paper flattening function; detecting the paper lighting situation through a sensor, and automatically adjusting the lighting intensity and angle, specifically adjusted through the following formula: Wherein, I adj is the lighting intensity adjustment value, k is the adjustment coefficient, set according to the device characteristics and experience, I t is the target lighting intensity (the ideal lighting intensity on the paper surface, set according to the paper type and scanning requirements), I o is the currently detected lighting intensity on the paper surface, I maxIt is the maximum illumination intensity that the lighting device can provide. According to the difference between the current light and the target light, the illumination intensity is adjusted proportionally according to this formula to ensure that the paper receives appropriate light; at the same time, negative pressure adsorption or a flexible flattening device is used to reduce paper wrinkles. Three modes of high-resolution optical scanning, infrared scanning, and ultraviolet scanning are adopted. Among them, high-resolution optical scanning obtains the basic content of the document, infrared scanning detects anti-counterfeiting marks or watermarks hidden inside the paper, and ultraviolet scanning is for documents containing fluorescent anti-counterfeiting materials. After collection, the image data of the three modes are fused.

[0073] During the collection process, the image quality is monitored in real time through an image sharpness evaluation algorithm and a noise level detection method; among them, the image sharpness evaluation algorithm calculates the sharpness index C of the image through the variance method, where M and N are the number of rows and columns of the image respectively, and f(i, j) is the pixel value of the image at the coordinate (i, j), is the average pixel value of the image; the larger the sharpness index C, the clearer the image; a sharpness threshold C min (the specific value is determined according to experiments and experience) is set. When C < C min , it is determined that the image sharpness is insufficient and needs to be recollected; if the image sharpness is insufficient, the image of this page of the document is automatically recollected, and at most three attempts are made;

[0074] The noise level detection first calculates the difference between the image and the image after smoothing processing, e(i, j) = f(i, j) - g(i, j), where f(i, j) is the pixel value of the original image and g(i, j) is the pixel value of the smoothed image; then the noise level N is calculated, A noise threshold N max (specifically determined according to the characteristics and requirements of the image sensor) is set. When N > N max , it is considered that there is too much noise and the image needs to be recollected.

[0075] Image preprocessing: For the problems of noise, contrast, and information loss in the collected image, corresponding algorithms are used for denoising, enhancing contrast, and repairing information respectively to optimize the image quality;

[0076] For the salt-and-pepper noise caused by paper texture and stains in the collected image, the median filtering algorithm is used; for Gaussian noise caused by uneven illumination or electronic device interference, the Gaussian filtering algorithm is adopted, and the filtering parameters are automatically adjusted according to the spatial distribution characteristics of the noise; the adaptive histogram equalization algorithm is used to enhance the contrast of the image to make the text and patterns clearer; for the information loss caused by paper damage or stains, it is repaired through interpolation algorithms and texture synthesis algorithms based on neighborhood pixel information, and appropriate algorithm parameters are selected according to the size of the missing area and the characteristics of surrounding pixels.

[0077] Feature extraction: On the basis of extracting conventional features, logical structure analysis is added to refine the text features and accurately extract the anti-counterfeiting element features, forming a rich and accurate document feature vector;

[0078] In addition to extracting traditional text, seal, and anti-counterfeiting logo features, analyze the logical structure of the document content, such as the relative position relationship and arrangement order between different information blocks; refine the text feature extraction, including the text tilt angle and stroke thickness change; the text tilt angle is calculated by calculating the tilt angle of the minimum bounding rectangle of the text area to determine the text tilt angle. Specifically, let the pixel coordinates of the text area be (x i , y i ), i = 1, 2,..., n, where n is the number of pixel points in the text area; first calculate the centroid coordinates (x c , y c ) of the text area: Then calculate the covariance matrix C:

[0079] Finally, by calculating the eigenvalues and eigenvectors of the covariance matrix, obtain the tilt angle θ of the minimum bounding rectangle: where C 11 , C 12 , C 22 are the elements of the covariance matrix C; this angle is used as the text tilt feature for subsequent matching and analysis;

[0080] For the special anti-counterfeiting elements such as microtext and rainbow printing patterns in paper real estate certificate documents, use high-resolution images and specialized image analysis algorithms to accurately extract their features; analyze the size, spacing, arrangement method of microtext and the color transition feature of rainbow printing patterns; for the size and spacing features of microtext, quantify them by calculating the pixel statistical information of the microtext area. Let the pixel value of the microtext area be m(i, j), i = 1, 2,..., M w , j = 1, 2,..., N w , where M w and N w are the number of rows and columns of the microtext area; the calculation of the average size S (in pixels) of microtext is: where n w is the number of valid pixels (i.e., pixels belonging to the text part) in the microtext area, which is judged by setting a pixel value threshold;

[0081] The microtext spacing D (in pixels) is obtained by calculating the average distance between adjacent microtext areas; first detect the boundaries of the microtext areas, then calculate the distance between adjacent boundaries, and finally take the average value to get the spacing D; these quantified features more accurately describe the microtext for anti-counterfeiting verification.

[0082] Information matching and verification: Compare the characteristics of the document to be verified with those of the same type and the standard template, statistically analyze the distribution law of similar information, use a machine learning model to identify error types, and dynamically calculate the similarity threshold for accurate judgment;

[0083] Compare the characteristics of the document to be verified with those of the property certificate documents of the same type and the standard template. By statistically analyzing the distribution law of similar information in documents of the same type, including the property area range in the same region and the format of the property owner's name, preliminarily screen suspicious information; Specifically:

[0084] Collect a large amount of existing and verified information on property certificate documents of the same type, including various relevant information such as property area, property owner's name, property location, and house structure, and store it in the database in a certain format and structure to ensure the accuracy and integrity of the data. At the same time, classify and sort the data, for example, classify it according to dimensions such as region and property type (residential, commercial, etc.); Extract the characteristics of the document to be verified through the feature extraction step and form a complete feature vector of the document to be verified; For the property area: Calculate the mean and standard deviation of the property areas of the same type, determine a reasonable range, and screen suspicious information; For the property owner's name: Screen relevant records from the database of property information of the same type, extract the property owner's name information; Statistically analyze all property owner surnames, calculate the frequency of each surname; According to the statistical results, set the surname frequency threshold; Extract the property owner surnames in the document to be verified and judge whether their frequency is lower than the threshold. If it is lower than the threshold, there may be an abnormality. Mark the document as suspicious in the property owner's name format and record relevant information (such as document number, property owner's name, the frequency of this surname in documents of the same type, etc.), and subsequent combined with other information can further judge whether there is a problem;

[0085] Use a machine learning algorithm trained based on a large amount of verified document data to build a model to identify common error types in scanning and recognition, including error patterns such as similar character confusion ("0" and "O", "1" and "l"), information duplication or omission, etc., to assist in judging whether there are such low-level errors;

[0086] When calculating the similarity between the document to be verified, the standard template, and the real estate information of the same type, dynamically adjust the similarity thresholds of different features according to document type, age, and regional factors; specifically, after obtaining the document type, age, and regional factors, evaluate the document type, age, and regional factors respectively to obtain the type score, age score, and regional score. After normalization, establish a bottom circle with the sum of the type score and the age score as the radius, and establish a cone model with the regional score as the height. Calculate the volume of the cone model and record it as the threshold judgment value. Compare the obtained threshold judgment value with several preset threshold judgment value intervals. Each of the several preset threshold judgment value intervals corresponds to a different similarity threshold. When determining the threshold judgment value interval to which the threshold judgment value belongs, determine the similarity threshold and substitute this similarity threshold into the calculation process of the similarity between the document to be verified, the standard template, and the real estate information of the same type.

[0087] Result output and report generation: Generate a verification report containing multi-faceted information, output it in multiple formats, mark and record suspicious information areas, send the report to human reviewers and receive feedback for processing to complete the verification process;

[0088] Generate a paper real estate certificate document information verification report based on the information matching and verification results. The report content includes basic document information, detailed image acquisition information (acquisition mode, quality assessment results, number of re-acquisitions, etc.), description of the preprocessing process, feature extraction results, comparison with the standard template and real estate information of the same type, details of the matching degree, and the final verification conclusion. Present the report in an intuitive and easy-to-understand way, such as showing the change of feature similarity through charts and marking suspicious information areas with images, and support multiple format outputs (such as PDF, HTML, XML, etc.) for easy archiving and retrieval; The processing process for suspicious information areas is as follows:

[0089] Determine the annotation specifications for suspicious information areas, including the symbols, colors, shapes for clear annotation, and format requirements for the annotation content; classify the possible suspicious information, such as seal types (including blurred seals, seals with shapes inconsistent with standards, abnormal seal colors, etc.), text information types (such as blurred text, confusion of similar characters, information repetition or omission, significant differences between key information and standard templates, etc.), overall image types (such as low image clarity, presence of shadows or occlusions that affect information recognition, etc.). For different types of suspicious information, formulate corresponding annotation methods and prompt messages so that manual review personnel can quickly understand the location of the suspicious points; use image processing software or develop specialized annotation tools with functions such as drawing rectangular frames and adding text annotations on images. Ensure that the annotation tools can be integrated with the real estate information verification system for convenient direct invocation during the report generation process; when generating a report, according to the suspicious information areas determined during the information matching and verification process, use the annotation tools to annotate according to the established specifications. For suspicious seal areas, draw accurate rectangular frames based on the position and scope of the seal, and annotate "suspicious seal" and the specific reasons; for suspicious text information areas, if similar character confusion is found, select the confused characters and annotate detailed information such as "similar characters, may be confused (such as '0' and 'O')"; for overall image problems, annotate "image quality problem" and relevant descriptions at the edge or blank area of the image;

[0090] Meanwhile, establish a record form for suspicious area information, recording the location of each suspicious area (such as the upper left corner coordinates and lower right corner coordinates of the rectangular box), suspicious type, relevant characteristic values (such as text feature vector values, similarity values with the standard template, etc.), and the corresponding file page numbers (if it is a multi-page document). These recorded information, together with the labeled images, will provide detailed data support for manual review; send the generated report to the manual review personnel through the task assignment mechanism within the system or by email, etc. When sending an email, clearly mark "Real Estate Certificate Document Verification Report - Suspicious Documents Require Review" in the email subject, and briefly explain the importance of the report and the key points for review in the body of the email, reminding the review personnel to focus on the suspicious information areas. At the same time, provide a link to view the report (if it is an online viewing format) or an attachment download method (if it is in formats such as PDF), ensuring that the review personnel can easily obtain the report; after receiving the report, the review personnel, in accordance with the specifications and training requirements, carefully view the content of the report, focusing on the labeled suspicious information areas. Combining their professional knowledge and experience, conduct in-depth analysis of the suspicious areas, such as comparing with the real samples of the seals, consulting relevant real estate file materials to verify the text information, etc. During the review process, use the review tool to record their review opinions and discovered problems, such as confirming whether there are indeed problems in the suspicious areas, the severity of the problems, and possible cause analysis, etc.; after the review personnel complete the review, feedback the review results to the real estate information verification system. The review results should include the final judgment on the suspicious areas (confirming problems, ruling out suspicion, or requiring further investigation), handling suggestions (such as requesting supplementary materials, re-scanning the document, rejecting the document, etc.), and improvement suggestions for the verification system (such as improving the labeling accuracy, increasing the dimension of feature extraction, etc.). The system updates the verification status of the document according to the review results and performs corresponding subsequent operations according to the handling suggestions, such as notifying the applicant to supplement materials or marking and archiving the documents with serious problems for subsequent query and statistical analysis.

[0091] A real estate information verification system based on image technology, comprising:

[0092] The image acquisition module is composed of a device control unit, a multi-mode scanning unit, and an image quality monitoring unit;

[0093] The device control unit controls the acquisition device, adjusts the lighting according to the lighting conditions, reduces paper wrinkles, and ensures the acquisition of high-quality images; the multi-mode scanning unit executes three scanning modes and fuses the data to obtain comprehensive document information; the image quality monitoring unit monitors the image quality in real time, judges according to the threshold and triggers re-acquisition to ensure that the images are qualified.

[0094] The image preprocessing module is composed of a noise removal unit, an image enhancement unit, and an information repair unit;

[0095] The noise removal unit adopts corresponding filtering algorithms and adjusts parameters for different noises to remove the noises; the image enhancement unit uses an algorithm to enhance the image contrast for subsequent processing; the information repair unit repairs the information loss caused by damage or stains;

[0096] The feature extraction module consists of a conventional feature extraction unit, a logical structure analysis unit, a text refinement feature unit, and an anti-counterfeiting element feature unit;

[0097] The conventional feature extraction unit is used to extract basic features such as text and seals; the logical structure analysis unit is used to analyze the logical structure of the document content to assist in judgment; the text refinement feature unit is used to refine the text features and accurately depict the text; the anti-counterfeiting element feature unit extracts the features of special anti-counterfeiting elements for anti-counterfeiting verification;

[0098] The information matching and verification module includes a same-type data comparison unit, a machine learning model unit, and a similarity calculation unit;

[0099] The same-type data comparison unit compares same-type data and screens out suspicious information such as the property area and the name of the property owner; the machine learning model unit uses the model to identify common types of scanning errors; the similarity calculation unit dynamically calculates the similarity threshold to accurately judge the document matching degree;

[0100] The result output and report generation module includes a report generation unit, a suspicious information marking unit, and a report sending and feedback unit;

[0101] The report generation unit generates a verification report containing multi-faceted information and outputs it in multiple formats; the suspicious information marking unit marks the suspicious areas and records the relevant information to assist manual review; the report sending and feedback unit sends the report to the manual reviewer, receives the feedback and processes it to improve the verification process.

[0102] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not elaborate on all the details, nor do they limit the invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the relevant technical fields can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for verifying real estate information based on image technology, characterized in that, including; S1. Image acquisition: Using a collection device equipped with multiple functions, through adjustable lighting, automatic paper flattening, multi-mode scanning, and real-time quality monitoring means, obtain high-quality and comprehensive images of paper real estate certificate documents; S2. Image preprocessing: For the problems of noise, contrast, and information loss in the collected images, use corresponding algorithms to perform denoising, contrast enhancement, and information repair operations respectively to optimize the image quality; S3. Feature extraction: On the basis of extracting conventional features, add logical structure analysis, refine text features, and extract anti-counterfeiting element features to form rich and accurate document feature vectors; S4. Information matching and verification: Compare the features of the document to be verified with those of the same type and standard template, statistically analyze the distribution law of similar information, and use a machine learning model to identify error types. At the same time, dynamically calculate the similarity threshold for accurate judgment; S5. Result output and report generation: Generate a verification report containing various aspects of information, output it in multiple formats, mark and record suspicious information areas, send the report to manual review personnel and receive feedback for processing to complete the verification process.

2. The real estate information verification method based on image technology according to claim 1, wherein The specific steps of S1 are as follows: Collect images of paper real estate certificate documents using a collection device equipped with an adjustable lighting system and an automatic paper leveling function; detect the paper lighting conditions through sensors and automatically adjust the lighting intensity and angle, and the adjustment is specifically carried out through the following formula: Where, I adj is the lighting intensity adjustment value, k is the adjustment coefficient, set according to the device characteristics and experience, I t is the target lighting intensity, I o is the lighting intensity detected on the current paper surface, I max is the maximum lighting intensity provided by the lighting device. According to this formula, the lighting intensity is adjusted proportionally according to the difference between the current lighting and the target lighting to ensure that the paper receives the correct lighting; At the same time, use a negative pressure adsorption or flexible flattening device to reduce paper wrinkles; adopt three scanning modes: optical scanning, infrared scanning, and ultraviolet scanning. Among them, optical scanning obtains the basic content of the document, infrared scanning detects hidden anti-counterfeiting marks or watermarks inside the paper, and ultraviolet scanning is for documents containing fluorescent anti-counterfeiting materials. After collection, fuse the image data of the three modes; During the acquisition process, the image quality is monitored in real time through an image sharpness evaluation algorithm and a noise level detection method; among them, the image sharpness evaluation algorithm calculates the sharpness index C of the image through the variance method, where M and N are the number of rows and columns of the image respectively, and f(i, j) is the pixel value of the image at the coordinate (i, j), and is the average pixel value of the image; the larger the sharpness index C, the clearer the image; a sharpness threshold C min is set. When C < C min , it is determined that the image sharpness is insufficient and needs to be re-acquired; if the image sharpness is insufficient, the image of this page of the document is automatically re-acquired. The noise level detection first calculates the difference between the image and the smoothed image, e(i, j) = f(i, j) - g(i, j), where f(i, j) is the pixel value of the original image and g(i, j) is the pixel value of the smoothed image; then calculates the noise level N. Set the noise threshold N max , when N > N max , it is determined that there is too much noise and the image needs to be recollected.

3. A method for verifying real estate information based on image technology according to claim 1, wherein, The specific operation steps of S2 are as follows: For the salt-and-pepper noise caused by paper texture and stains in the collected images, use the median filtering algorithm; for the Gaussian noise generated by uneven lighting or electronic device interference, adopt the Gaussian filtering algorithm, and automatically adjust the filtering parameters according to the spatial distribution characteristics of the noise; Use the adaptive histogram equalization algorithm to enhance the image contrast to make the text and patterns clearer; for the information loss caused by paper damage or stains, repair it through the interpolation algorithm and texture synthesis algorithm based on neighborhood pixel information, and select the corresponding algorithm parameters according to the size of the missing area and the characteristics of surrounding pixels.

4. A method for verifying real estate information based on image technology according to claim 1, characterized in that, The specific operation steps of S3 are as follows: In addition to extracting the features of text, seals, and anti-counterfeiting marks, add an analysis of the logical structure of the document content, including the relative position relationship and arrangement order between different information blocks; refine the text feature extraction, including the text tilt angle and the change in stroke thickness; For the anti-counterfeiting elements such as microtext and rainbow printing patterns in paper real estate certificate documents, use image and image analysis algorithms to extract their features; Analyze the size, spacing, arrangement of the microtext, and the color transition characteristics of the rainbow printing pattern; for the size and spacing characteristics of the microtext, quantify them by calculating the pixel statistical information of the microtext area. Let the pixel value of the microtext area be m(i, j), where i = 1, 2, …, M w , j = 1, 2, …, N w , where M w and N w are the number of rows and columns of the microtext area; the calculation of the average size S of the microtext is as follows: where n w is the number of valid pixels in the microtext area, which is judged by setting a pixel value threshold; The microtext spacing D is obtained by calculating the average distance between adjacent microtext areas; first, detect the boundaries of the microtext areas, then calculate the distance between adjacent boundaries, and finally calculate the average value to obtain the spacing D; these quantified features more accurately describe the microtext for anti-counterfeiting verification.

5. The method for verifying real estate information based on image technology according to claim 4, wherein, The specific operation steps for calculating the text tilt angle in S3 are as follows: The text skew angle is calculated by calculating the skew angle of the minimum bounding rectangle of the text area. Specifically, let the pixel coordinates of the text area be (x i , y i ), where i = 1, 2, …, n and n is the number of pixel points in the text area. First, calculate the centroid coordinates (x c , y c ) of the text area: Then calculate the covariance matrix C: Finally, by calculating the eigenvalues and eigenvectors of the covariance matrix, the inclination angle θ of the minimum circumscribed rectangle is obtained: where C 11 , C 12 , C 22 are the elements of the covariance matrix C; this angle is used as a text inclination feature for subsequent matching and analysis.

6. The real estate information verification method based on image technology according to claim 1, wherein The specific operation steps of S4 are as follows: Compare the characteristics of the document to be verified with those of real estate certificate documents of the same type and the standard template. By statistically analyzing the distribution patterns of similar information in documents of the same type, including the real estate area range in the same region and the format of the property owner's name, preliminarily screen for suspicious information. Specifically: Collect information on existing and verified real estate certificate documents of the same type, including real estate area, property owner's name, real estate location, and information related to the house structure. Store this information in a database according to a preset format and structure to ensure the accuracy and integrity of the data. At the same time, classify and organize the data according to dimensions such as region and real estate type. Extract the characteristics of the document to be verified through the feature extraction step and form a complete feature vector of the document to be verified. Regarding the real estate area: Calculate the mean and standard deviation of the real estate areas of the same type, determine a reasonable range, and screen for suspicious information. Regarding the property owner's name: Screen relevant records from the database of real estate information of the same type, extract the property owner's name information; count all the surnames of the property owners, calculate the frequency of each surname; according to the statistical results, set a surname frequency threshold; extract the surnames of the property owners in the document to be verified and judge whether their frequency is lower than this threshold. If it is lower than the threshold, judge that there is an abnormal situation, mark this document as suspicious in the property owner's name format, and record relevant information, including the document number, the property owner's name, and the frequency of this surname in documents of the same type. Use a machine learning algorithm trained based on the data of verified documents to build a model to identify common error types in scanning and recognition, including similar character confusion, information repetition or missing error patterns, and assist in judging whether there are such errors. When calculating the similarity between the document to be verified and the standard template and real estate information of the same type, dynamically adjust the similarity thresholds of different characteristics according to factors such as document type, era, and region.

7. The method for verifying real estate information based on image technology according to claim 6, wherein The specific operation steps for adjusting the similarity thresholds of different characteristics in S4 are as follows: After obtaining the document type, era, and region factors, evaluate the document type, era, and region factors respectively to obtain a type score, an era score, and a region score. After normalization, use the sum of the type score and the era score as the radius to establish a bottom circle, use the region score as the height to establish a cone model, calculate the volume of this cone model, and record it as the threshold judgment value. Compare the obtained threshold judgment value with several preset threshold judgment value intervals. Each of the several preset threshold judgment value intervals corresponds to a different similarity threshold. When determining the threshold judgment value interval to which the threshold judgment value belongs, determine the similarity threshold and substitute this similarity threshold into the calculation process of the similarity between the document to be verified and the standard template and real estate information of the same type.

8. A method for verifying real estate information based on image technology according to claim 1, characterized in that, The specific operation steps of S5 are as follows: Generate a paper real estate certificate information verification report based on the information matching and verification results. The report content includes the basic information of the document, the detailed situation of image acquisition, specifically the acquisition mode, quality assessment results, number of re-acquisitions, description of the preprocessing process, feature extraction results, comparison with the standard template and the same type of real estate information, matching degree details, and the final verification conclusion; present the report, including showing the change of feature similarity through charts and marking the suspicious information areas with images, and support multiple format outputs, including PDF, HTML, XML, for archiving and reference.

9. The method for verifying real estate information based on image technology according to claim 8, wherein The processing process for the suspicious information areas in S5 is as follows: Determine the annotation specifications for the suspicious information areas, including clarifying the symbols, colors, shapes for annotation, and the format requirements for the annotation content; classify the emerging suspicious information, including seal types, text information types, and overall image types; for different types of suspicious information, formulate corresponding annotation methods and prompt information; Use image processing software or annotation tools to draw rectangular frames on the image and add text annotation functions; When generating the report, according to the suspicious information areas determined during the information matching and verification process, use the annotation tools to mark according to the established specifications; At the same time, establish a suspicious area information record form to record the location, suspicious type, relevant feature values of each suspicious area, and the corresponding document page numbers; these recorded information will provide detailed data support for manual review together with the marked images; Send the generated report to the manual review personnel through the task allocation mechanism within the system or by email; when sending the email, mark it in the email subject and explain the importance of the report and the key points for review in the body of the email, reminding the review personnel to focus on the suspicious information areas; at the same time, provide the link for viewing the report or the attachment download method to ensure that the review personnel can conveniently obtain the report; After receiving the report, the review personnel, according to the specifications, view the report content and focus on the marked suspicious information areas; combine their own professional knowledge and experience to conduct in-depth analysis on the suspicious areas; during the review process, use the review tools to record their review opinions and the problems found; After the review personnel complete the review, feedback the review results; the review results include the final judgment on the suspicious areas, processing suggestions, and improvement suggestions for the verification system; Update the verification status of the document according to the review results and perform corresponding subsequent operations according to the processing suggestions.

10. A system applied to a method for verifying real estate information based on image technology according to any one of claims 1-9, characterized in that, Including: An image acquisition module, which consists of a device control unit, a multi-mode scanning unit, and an image quality monitoring unit; The device control unit controls the acquisition device, adjusts the illumination according to the light conditions, reduces paper wrinkles, and ensures high-quality image acquisition; the multi-mode scanning unit executes three scanning modes and fuses the data to obtain comprehensive document information; the image quality monitoring unit monitors the image quality in real time, judges according to the threshold and triggers re-acquisition to ensure the image is qualified; An image preprocessing module, which consists of a noise removal unit, an image enhancement unit, and an information repair unit; The noise removal unit adopts corresponding filtering algorithms and adjusts parameters for different noises to remove the noises; The image enhancement unit uses algorithms to enhance the image contrast for subsequent processing; the information repair unit repairs the information loss caused by damage or stains; The feature extraction module consists of a conventional feature extraction unit, a logical structure analysis unit, a text refinement feature unit, and an anti-counterfeiting element feature unit; The conventional feature extraction unit is used to extract the basic features of text and seals; The logical structure analysis unit is used to analyze the logical structure of the document content for auxiliary judgment; the text refinement feature unit is used to refine the text features and accurately depict the text; the anti-counterfeiting element feature unit extracts anti-counterfeiting element features for anti-counterfeiting verification; The information matching and verification module consists of a same-type data comparison unit, a machine learning model unit, and a similarity calculation unit; The same-type data comparison unit is used to compare same-type data and screen for suspicious information such as the property area and the name of the property owner; The machine learning model unit uses the model to identify common types of scanning errors; the similarity calculation unit dynamically calculates the similarity threshold to accurately judge the document matching degree; The result output and report generation module includes a report generation unit, a suspicious information annotation unit, and a report sending and feedback unit; The report generation unit generates a verification report containing multi-faceted information and outputs it in multiple formats; the suspicious information annotation unit is used to mark the suspicious areas and record the relevant information to assist manual review; the report sending and feedback unit sends the report to the manual review personnel, receives the feedback and processes it to improve the verification process.

Images