Control system for urban data governance

By combining natural language processing and computer vision technologies, the system identifies text and image data uploaded by users, solving the problem of low efficiency in identifying illegal parking in traditional methods. It achieves real-time response and optimization of monitoring resources, thereby improving the efficiency of urban traffic management and public participation.

CN119048990AInactive Publication Date: 2024-11-29FUJIAN BIG DATA GROUP LONGYAN CO LTD

Patent Information

Application Number
CN202411525812.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-30
Publication Date
2024-11-29
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing technologies struggle to efficiently and cost-effectively identify information related to illegal parking from massive amounts of user-uploaded data. Traditional methods are inefficient and lack real-time response capabilities.

Method used

By employing natural language processing and computer vision technologies, combined with data acquisition, text processing, and image processing modules, the system identifies and extracts illegal parking features from user-uploaded text descriptions and image data. The data analysis module then statistically analyzes and marks areas with insufficient monitoring.

Benefits of technology

It has enabled efficient and accurate identification and real-time response to illegal parking information, optimized the allocation of urban monitoring resources, and improved urban traffic order and public participation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119048990A_ABST
    Figure CN119048990A_ABST
Patent Text Reader

Abstract

The invention discloses a control system for urban data governance, and belongs to the technical field of data governance. Collecting character description and image data uploaded by the user; identifying the text description through a natural language processing technology, and screening the text description containing illegal parking keywords; preprocessing the image data, and extracting illegal parking features from the preprocessed image data by using a computer vision technology; counting license plate numbers and parking positions corresponding to illegal parking events according to the effective character description and the effective image data, obtaining roads related to all the parking positions, counting the proportion of the number of the illegal parking events of each related road in the total number, and obtaining the proportion of the number of monitoring cameras of each related road in the total number; marking the two related roads with relatively large proportion difference values as roads with insufficient monitoring; according to the invention, by collecting, analyzing and utilizing the uploaded data of a large number of users, decision support is provided for urban management departments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data governance, and in particular to a control system for urban data governance. Background Art

[0002] Urban data governance refers to the collection, organization, analysis and utilization of large amounts of data generated during urban operations through systematic methods and technical means to achieve intelligent urban management and optimized decision-making.

[0003] With the acceleration of urbanization, urban traffic problems are becoming increasingly prominent. Illegal parking, as a major problem in urban management, not only affects the city's traffic order, but also has a negative impact on the city's appearance. Traditional urban traffic data management methods mainly rely on manual inspections and surveillance cameras, but this method has the problems of low efficiency, high cost and difficulty in real-time response.

[0004] In recent years, with the development of Internet technology, more and more users have uploaded text descriptions and image data about illegal parking in cities through social media, mobile applications and other platforms. These data contain rich information. If they can be effectively collected and analyzed, they will provide new ideas and methods for urban data governance.

[0005] Therefore, how to accurately identify information related to illegal parking from massive amounts of user-uploaded data and further extract useful information has become a problem that needs to be solved urgently. Summary of the invention

[0006] The purpose of the present invention is to provide a control system for urban data management to solve the following technical problems: How to accurately identify information related to illegal parking from massive amounts of user-uploaded data as a supplement to monitoring data.

[0007] The purpose of the present invention can be achieved through the following technical solutions: A control system for urban data governance, comprising: Data collection module, used to collect text descriptions and image data uploaded by users; a text processing module, for identifying the text descriptions by natural language processing technology, screening the text descriptions containing the keyword of illegal parking, marking them as target text descriptions, identifying whether the target text descriptions contain location information and / or license plate numbers, and if so, extracting the location information and / or license plate numbers, and marking the target text descriptions as valid text descriptions; An image processing module, used to pre-process the image data, extract illegal parking features from the pre-processed image data using computer vision technology, wherein the features include vehicle outline, license plate number, and parking position, and mark the image data containing the features as valid image data; The data analysis module is used to count the license plate numbers and parking positions corresponding to illegal parking events based on the valid text descriptions and valid image data, obtain the roads involved in all parking positions, count the number of illegal parking events on each involved road and the proportion Pi of the total number, and obtain the number of surveillance cameras on each involved road and the proportion Ki of the total number, calculate the Pi-Ki ratio difference of each road, and mark the involved roads with a ratio difference greater than a preset threshold as insufficiently monitored roads.

[0008] As a further solution of the present invention: the data collection module collects the data uploaded by the user based on the crawler technology by setting the monitoring website and the monitoring content range.

[0009] As a further solution of the present invention: the process of the text processing module recognizing the text description is as follows: The text description is filtered, punctuation marks and stop words in the text description are removed, the filtered text description is segmented into a plurality of words, and the words are converted into vectors based on a bag-of-words model and TF-IDF to extract features; a text classifier is constructed based on a machine learning algorithm, the text classifier is trained using a labeled data set, and the accuracy of the text classifier is verified based on cross-validation; the word vector is input into the verified text classifier, the context feature information of any word vector is deeply learned, and keywords related to illegal parking are output; and the location information and / or license plate number in the text description are recognized through a place name dictionary and a digital format.

[0010] As a further solution of the present invention: in the image processing module, the process of preprocessing the image data is: The pixel size of the image data is adjusted to a set size, the image data is subjected to noise reduction processing based on a Gaussian filter, and the contrast is adjusted to increase the details and features in the image data.

[0011] As a further solution of the present invention: the process of extracting features by the image processing module is: A feature object recognition model is constructed based on a pre-trained Transformer network. The feature object recognition model is trained through labeled supervised learning. Artificial objects in image data are recognized through an edge detection algorithm. The artificial objects include vehicles, roads, license plates and road signs. Corner points of artificial objects are recognized through corner point detection. Artificial objects, pedestrians and landscape plants are distinguished based on texture analysis. The image is divided into several areas, each area corresponds to a feature object. Any area is input into the trained feature object recognition model, and the recognition result is output. The license plate information and road position information in the license plate and road sign are recognized based on OCR text recognition technology.

[0012] As a further solution of the present invention: in the image processing module, when the road location information cannot be identified from the image data, artificial objects whose proportion in the image data is greater than a preset proportion are obtained, and similarity comparison is performed on the artificial objects through a landmark building image library to identify whether the artificial objects are landmark buildings. If so, the location information of the landmark building is obtained.

[0013] As a further solution of the present invention: the data acquisition module is further used to obtain the upload time of the user uploading the valid text description and the valid image data, and when the user upload time is obtained through the traffic big data, the congestion index of the corresponding road involved, the congestion index includes smooth, slow, medium congestion and severe congestion; In the data analysis module, for any road involved, when a user uploads valid text description and valid image data related to the road involved, and the congestion index of the road involved is moderate congestion or severe congestion, it is fed back that the cause of congestion on the road involved is illegal parking.

[0014] As a further solution of the present invention: in the image processing module, the vehicle outline includes but is not limited to motor vehicles, non-motor vehicles and mobile stall vehicles.

[0015] Beneficial effects of the present invention: (1) By combining natural language processing technology and computer vision technology, the present invention can accurately identify information related to illegal parking from text descriptions and image data uploaded by users. By using the data uploaded by users for real-time analysis, the urban management department can quickly respond to illegal parking incidents and take corresponding management measures. This real-time and dynamic feature helps to solve traffic problems in a timely manner and reduce the impact of illegal parking on urban traffic order. By collecting, analyzing and using a large amount of user-uploaded data, it provides decision-making support for urban management departments. This data-driven approach helps to formulate urban management strategies in a more scientific and accurate manner. (2) By statistically analyzing the number of illegal parking incidents and the number of surveillance cameras on different roads, the ratio difference is calculated, and roads with a ratio difference greater than a preset threshold are marked as under-surveilled roads; this helps urban management departments to rationally allocate surveillance resources and ensure that key areas are adequately monitored; (3) The present invention can encourage users to upload text descriptions and image data about illegal parking, thereby increasing the channels for public participation in urban management; this not only increases the public's attention and satisfaction with urban management, but also promotes interaction and cooperation between the government and the public. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] See also Figure 1 As shown, the present invention is a control system for urban data management, comprising: Data collection module, used to collect text descriptions and image data uploaded by users; a text processing module, for identifying the text descriptions by natural language processing technology, screening the text descriptions containing the keyword of illegal parking, marking them as target text descriptions, identifying whether the target text descriptions contain location information and / or license plate numbers, and if so, extracting the location information and / or license plate numbers, and marking the target text descriptions as valid text descriptions; An image processing module, used to pre-process the image data, extract illegal parking features from the pre-processed image data using computer vision technology, wherein the features include vehicle outline, license plate number, and parking position, and mark the image data containing the features as valid image data; The data analysis module is used to count the license plate numbers and parking positions corresponding to illegal parking events based on the valid text descriptions and valid image data, obtain the roads involved in all parking positions, count the number of illegal parking events on each involved road and the proportion Pi of the total number, and obtain the number of surveillance cameras on each involved road and the proportion Ki of the total number, calculate the Pi-Ki ratio difference of each road, and mark the involved roads with a ratio difference greater than a preset threshold as insufficiently monitored roads.

[0020] By combining natural language processing technology and computer vision technology, illegal parking incidents can be efficiently and accurately identified from text descriptions and image data uploaded by users. The system not only improves the accuracy and efficiency of illegal parking identification, realizes real-time response and dynamic management, but also provides data support for the optimal allocation of urban monitoring resources, thereby promoting the improvement of urban traffic order and the improvement of residents' quality of life.

[0021] In another preferred embodiment of the present invention, the data collection module collects the data uploaded by the user based on crawler technology by setting the monitoring website and the monitoring content range.

[0022] In another preferred embodiment of the present invention, the process of the text processing module recognizing the text description is as follows: The text description is filtered, punctuation marks and stop words in the text description are removed, the filtered text description is segmented into a plurality of words, and the words are converted into vectors based on a bag-of-words model and TF-IDF to extract features; a text classifier is constructed based on a machine learning algorithm, the text classifier is trained using a labeled data set, and the accuracy of the text classifier is verified based on cross-validation; the word vector is input into the verified text classifier, the context feature information of any word vector is deeply learned, and keywords related to illegal parking are output; and the location information and / or license plate number in the text description are recognized through a place name dictionary and a digital format.

[0023] By combining natural language processing technology, it is possible to efficiently and accurately identify information related to illegal parking from the text descriptions uploaded by users. This system not only improves the accuracy and efficiency of illegal parking identification, but also provides strong support for subsequent data integration and analysis.

[0024] In another preferred embodiment of the present invention, in the image processing module, the process of preprocessing the image data is: The pixel size of the image data is adjusted to a set size, the image data is subjected to noise reduction processing based on a Gaussian filter, and the contrast is adjusted to increase the details and features in the image data.

[0025] In another preferred embodiment of the present invention, the process of extracting features by the image processing module is: A feature object recognition model is constructed based on a pre-trained Transformer network. The feature object recognition model is trained through labeled supervised learning. Artificial objects in image data are recognized through an edge detection algorithm. The artificial objects include vehicles, roads, license plates and road signs. Corner points of artificial objects are recognized through corner point detection. Artificial objects, pedestrians and landscape plants are distinguished based on texture analysis. The image is divided into several areas, each area corresponds to a feature object. Any area is input into the trained feature object recognition model, and the recognition result is output. The license plate information and road position information in the license plate and road sign are recognized based on OCR text recognition technology.

[0026] By combining the pre-trained Transformer network in computer vision technology, the model can efficiently and accurately identify man-made objects related to illegal parking (such as vehicles, roads, license plates and road signs) from the image data uploaded by users, significantly improving the accuracy and efficiency of recognition. The feature object recognition model trained by labeled supervised learning can achieve real-time response and dynamic management of illegal parking incidents, solve traffic problems in a timely manner, and reduce the impact of illegal parking on urban traffic order. By identifying man-made objects in image data through edge detection algorithms, distinguishing between man-made objects, pedestrians and landscape plants based on texture analysis, and dividing the image into several areas, it helps urban management departments to reasonably allocate monitoring resources and ensure that key areas are adequately supervised.

[0027] In another preferred embodiment of the present invention, in the image processing module, when the road location information cannot be identified from the image data, artificial objects whose proportion in the image data is greater than a preset proportion are obtained, and similarity comparison is performed on the artificial objects through a landmark building image library to identify whether the artificial objects are landmark buildings. If so, the location information of the landmark building is obtained.

[0028] When the road location information cannot be identified from the image data, the artificial objects that account for more than a preset proportion in the image data are obtained, and the landmark image library is used for similarity comparison to identify whether the artificial object is a landmark building. If it is confirmed to be a landmark building, its location information can be obtained. This method provides an effective alternative to locate and identify roads, especially when road signs are not obvious or missing. It enhances the robustness and flexibility of the system, ensuring that illegal parking locations can be accurately identified and located in various complex scenarios.

[0029] In another preferred embodiment of the present invention, the data acquisition module not only obtains the valid text description and valid image data uploaded by the user, but also records the upload time of these data. Through the traffic big data platform, such as various map navigation apps, traffic broadcasts and traffic management platforms, the congestion index of the roads involved in these upload time points can be obtained in real time, including different states such as unimpeded, slow, moderately congested and severely congested.

[0030] In the data analysis module, for any road with moderate or severe congestion, if a user happens to upload valid text descriptions and valid image data about illegal parking on the road, it means that the congestion on the road is caused by illegal parking. Therefore, the system will feedback the cause of the congestion on the road to the traffic big data platform. This method can accurately identify and locate traffic congestion problems caused by illegal parking in real time, thereby providing timely and effective decision support for urban management departments.

[0031] In another preferred embodiment of the present invention, in the image processing module, the vehicle outline includes but is not limited to motor vehicles, non-motor vehicles and mobile stall vehicles.

[0032] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A control system for urban data management, characterized in that: include: Data collection module, used to collect text descriptions and image data uploaded by users; a text processing module, for identifying the text descriptions by natural language processing technology, screening the text descriptions containing the keyword of illegal parking, marking them as target text descriptions, identifying whether the target text descriptions contain location information and / or license plate numbers, and if so, extracting the location information and / or license plate numbers, and marking the target text descriptions as valid text descriptions; An image processing module, used to pre-process the image data, extract illegal parking features from the pre-processed image data using computer vision technology, wherein the features include vehicle outline, license plate number, and parking position, and mark the image data containing the features as valid image data; The data analysis module is used to count the license plate numbers and parking positions corresponding to illegal parking events based on the valid text descriptions and valid image data, obtain the roads involved in all parking positions, count the number of illegal parking events on each involved road and the proportion Pi of the total number, and obtain the number of surveillance cameras on each involved road and the proportion Ki of the total number, calculate the Pi-Ki ratio difference of each road, and mark the involved roads with a ratio difference greater than a preset threshold as insufficiently monitored roads.

2. A control system for urban data management according to claim 1, characterized in that: The data collection module collects the data uploaded by users based on crawler technology by setting the monitoring website and monitoring content range.

3. A control system for urban data management according to claim 1, characterized in that: The process of the text processing module recognizing the text description is as follows: Filtering the text description, removing punctuation marks and stop words in the text description, performing word segmentation on the filtered text description, dividing it into a number of words, and converting the words into vectors based on a bag-of-words model and TF-IDF to extract features; A text classifier is built based on a machine learning algorithm, trained using annotated datasets, and verified for accuracy based on cross-validation. Word vectors are input into the verified text classifier, and deep learning is performed on the contextual feature information of any word vector to output keywords related to illegal parking. Location information and / or license plate numbers in text descriptions are identified through place name dictionaries and digital formats.

4. A control system for urban data management according to claim 1, characterized in that: In the image processing module, the process of preprocessing the image data is as follows: The pixel size of the image data is adjusted to a set size, the image data is subjected to noise reduction processing based on a Gaussian filter, and the contrast is adjusted to increase the details and features in the image data.

5. A control system for urban data management according to claim 4, characterized in that: The process of extracting features by the image processing module is as follows: A feature object recognition model is constructed based on a pre-trained Transformer network. The feature object recognition model is trained through labeled supervised learning. Artificial objects in image data are recognized through an edge detection algorithm. The artificial objects include vehicles, roads, license plates and road signs. Corner points of artificial objects are recognized through corner point detection. Artificial objects, pedestrians and landscape plants are distinguished based on texture analysis. The image is divided into several areas, each area corresponds to a feature object. Any area is input into the trained feature object recognition model, and the recognition result is output. The license plate information and road position information in the license plate and road sign are recognized based on OCR text recognition technology.

6. A control system for urban data management according to claim 5, characterized in that: In the image processing module, when the road location information cannot be identified from the image data, artificial objects whose proportion in the image data is greater than a preset proportion are obtained, and similarity comparison is performed on the artificial objects through the landmark building image library to identify whether the artificial objects are landmark buildings. If so, the location information of the landmark building is obtained.

7. A control system for urban data management according to claim 1, characterized in that: The data acquisition module is also used to obtain the upload time of the valid text description and valid image data uploaded by the user, and obtain the congestion index of the road involved when the user upload time is obtained through traffic big data. The congestion index includes smooth, slow, medium congestion and severe congestion; In the data analysis module, for any road involved, when a user uploads valid text description and valid image data related to the road involved, and the congestion index of the road involved is moderate congestion or severe congestion, it is fed back that the cause of congestion on the road involved is illegal parking.

8. A control system for urban data management according to claim 1, characterized in that: In the image processing module, the vehicle outline includes but is not limited to motor vehicles, non-motor vehicles and mobile stall vehicles.

Citation Information

Patent Citations

  • Intelligent control system and method for illegal parking

    CN109584594A

  • Method and system for determining limited parking place

    CN111179578A

  • Network security data privacy protection management system

    CN118484829A

  • System for regulating penalty of illegal parking andmethod thereof

    KR1020060119079A

Cited By

  • Urban governance large model detection method and system fusing multi-modal data

    CN119495007A