Intelligent glare measurement system based on threshold increment

Abstract

The invention relates to an intelligent glare measurement system based on threshold increment, and the system comprises a data collection module which is used for collecting road field image data; the data processing module is used for carrying out brightness parameter extraction and position information extraction on the road field image data; the threshold value increment calculation module is used for calculating a threshold value increment by utilizing a preset threshold value increment calculation model according to the extracted brightness parameter and the position information; and the display and early warning module is used for displaying the threshold value increment and related data to a user and automatically triggering an alarm mechanism when the measurement result exceeds a preset threshold value. Through deep fusion of software and hardware and construction of a special data acquisition circuit and an embedded controller, automatic, real-time and closed-loop self-correction glare measurement is realized, data acquisition and processing precision can be improved, field operation risks and system overall cost are reduced, and the system has remarkable technical and economic advantages.

Description

Technical Field

[0001] The present invention relates to the fields of digital image processing, automated measurement and special hardware design, and in particular to an intelligent glare measurement system based on threshold increment. Background Art

[0002] Currently, traditional road lighting glare measurement mainly relies on illuminance meters, luminance meters, laser rangefinders, and manual point placement. These traditional methods have the following major drawbacks in practical applications:

[0003] The operation is cumbersome and time-consuming:

[0004] Traditional measurement methods typically require setting up multiple measurement points on-site, requiring operators to manually locate each measurement point, debug the instrument, and record data. This cumbersome process, requiring multiple manual calibrations and point placements, often results in hours of on-site measurement time. Furthermore, on-site operations pose significant safety risks in busy road environments.

[0005] Poor data repeatability and low accuracy:

[0006] Because traditional methods rely on human judgment and manual operation, data collection standards and instrument operation methods vary significantly across measurement points. These methods are susceptible to operator subjective factors, environmental noise, and interference from light reflections, resulting in poor repeatability and insufficient precision. Furthermore, traditional instruments themselves have limited response speed and accuracy under complex lighting conditions, further impacting the stability of measurement results.

[0007] High cost of instruments and equipment:

[0008] Traditional measurement methods often require the use of multiple specialized instruments, such as illuminance meters, luminance meters, tape measures, and laser rangefinders. The coordinated use of these instruments not only increases system complexity but also leads to high overall equipment procurement and maintenance costs, making them difficult to promote and apply in large-scale road lighting monitoring and management.

[0009] Measurement of complex impacts of on-site environment:

[0010] In real-world road environments, on-site measurement data is susceptible to dynamic changes due to factors such as uneven light distribution, interference from reflected light, and obstructions from vehicles and buildings. This is especially true at night or in adverse weather conditions, where traditional instruments lack measurement stability and responsiveness, making it difficult to fully reflect actual lighting conditions.

[0011] In recent years, with the rapid development of digital image processing, automated control, and industrial camera technology, the use of non-contact digital image acquisition and processing for glare measurement has become a trend. Using industrial cameras to capture real-time road scene images and automatically extracting brightness and position information using digital image processing algorithms can, to a certain extent, overcome the inconvenience of manual measurement points and the unstable measurement data associated with traditional methods. However, most automated glare measurement systems currently on the market still rely primarily on general-purpose computing platforms and off-the-shelf software algorithms for image processing. Their hardware typically utilizes common data acquisition interfaces and existing signal conditioning circuits, making it difficult to meet the requirements for stable and real-time data transmission in complex field environments.

[0012] Specifically, the existing automation system has obvious deficiencies in the following aspects:

[0013] Weak hardware anti-interference capabilities: The general data acquisition circuits and interfaces used in most systems cannot be effectively optimized for factors such as temperature, humidity, and electromagnetic interference in the field environment, resulting in signal noise during data transmission, affecting image quality and subsequent processing accuracy.

[0014] Insufficient real-time performance and stability: General hardware platforms have certain bottlenecks in high-speed data acquisition and real-time processing, making it difficult to ensure continuous and stable data output in complex road environments, especially in traffic congestion or rapidly changing lighting conditions.

[0015] Lack of dedicated closed-loop feedback control: In existing systems, although software algorithms can achieve certain image processing and parameter extraction, due to the lack of dedicated hardware feedback mechanisms, it is difficult to achieve automatic correction in long-term operation. System parameters are easily affected by environmental fluctuations and drift, which in turn leads to increased measurement errors.

[0016] Therefore, there is an urgent need to develop an intelligent glare measurement system that integrates dedicated hardware design and advanced image processing algorithms. The system should have dedicated hardware such as dedicated data acquisition and signal conditioning circuits, embedded closed-loop feedback control chips, etc., which can achieve high-precision and low-latency conversion of data collected by industrial cameras, and automatically adjust image acquisition and processing parameters through hardware-level real-time feedback mechanism, thereby greatly improving the overall stability, real-time performance and measurement accuracy of the system. Summary of the Invention

[0017] The purpose of the present invention is to provide an intelligent glare measurement system based on threshold increment. The system uses an industrial camera to capture road scene images, adopts digital image processing methods to extract brightness and position information in the images, constructs a threshold increment calculation model based on the relevant technical reports of the International Commission on Illumination (CIE), and realizes automated, real-time and high-precision glare measurement through real-time online parameter optimization and self-correction mechanisms, thereby solving the problems of low efficiency, cumbersome operation and large errors existing in traditional measurement methods.

[0018] To achieve the above object, the present invention provides the following solutions:

[0019] An intelligent glare measurement system based on threshold increment, comprising:

[0020] Data acquisition module, used for collecting road scene image data;

[0021] A data processing module, configured to extract brightness parameters and position information from the road scene image data;

[0022] A threshold increment calculation module is used to calculate a threshold increment value based on the extracted brightness parameter and position information using a preset threshold increment calculation model;

[0023] The display and warning module is used to display the threshold increment value and related data to the user, and automatically trigger the alarm mechanism when the threshold increment value exceeds the preset threshold.

[0024] Optionally, the system further comprises:

[0025] The self-correction module is used to monitor the deviation between the measured data and the preset standard. When the deviation exceeds the standard, the feedback control algorithm is automatically called to dynamically adjust the image data acquisition parameters and image processing parameters, while optimizing the threshold increment calculation model to form a closed-loop control.

[0026] Optionally, the data acquisition module realizes real-time acquisition of road scene image data through automatic exposure, automatic focus and parameter adjustment, and transmits the data to the data processing module;

[0027] The data acquisition module has a built-in preset parameter library and can automatically select acquisition mode and exposure parameters according to on-site conditions.

[0028] Optionally, the data processing module includes:

[0029] Image preprocessing unit: used for preprocessing the road scene image data;

[0030] The brightness parameter extraction unit is used to combine the optical characteristics and imaging principles of the camera to establish an improved brightness calibration method. The pixel values of each area in the preprocessed image data are mapped to actual brightness values. The multi-point regression or nonlinear fitting algorithm is used to calculate the average brightness Lav of the road surface and the local brightness data of each light source area.

[0031] The position information extraction unit is used to automatically identify the position of lamps and other key light sources and extract position coordinates and area information by performing image registration and feature point matching on the pre-processed image data using a binocular vision algorithm or in combination with a pre-calibrated monocular camera method.

[0032] Optionally, the image preprocessing unit performs noise suppression, contrast enhancement, distortion correction and region segmentation processing on the road scene image data. During the preprocessing process, a filtering algorithm is used to eliminate random noise, and image enhancement technology is used to highlight the target area.

[0033] Optionally, the threshold increment calculation module includes:

[0034] A first calculation unit is used to calculate the equivalent light curtain brightness Lv for each light source using a pre-built mathematical model, and to cumulatively calculate the total light curtain brightness Lv,total;

[0035] The second calculation unit is configured to calculate a threshold increment value by using a preset threshold increment calculation model.

[0036] Optionally, the expression of the threshold increment calculation model is:

[0037] TI=f(Lv,total,Lav)

[0038] Where TI is the threshold increment value and f is the mapping function.

[0039] Optionally, the self-correction module includes:

[0040] Error monitoring unit, used to monitor the deviation between the measured data and the preset standard, and determine whether there is system parameter drift or environmental influence through statistical analysis and data comparison;

[0041] Automatic correction unit, which is used to automatically call the feedback control algorithm when the error exceeds the limit, dynamically adjust the image data acquisition parameters and image data processing parameters, and optimize the threshold increment calculation model to form a closed-loop control;

[0042] The historical data recording unit is used to archive the parameter adjustment records and system operation logs generated during the calibration process.

[0043] The beneficial effects of the present invention are:

[0044] This invention utilizes automated data acquisition, precise digital image processing, threshold increment calculation based on international standards, and a real-time self-correction feedback mechanism to achieve a closed-loop system from image acquisition to data processing, calculation, and real-time feedback. This system not only significantly improves measurement accuracy and efficiency technically, but also demonstrates significant advantages in economic benefits, operational safety, and data management, providing an innovative and feasible solution for road lighting glare monitoring and traffic safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 Schematic diagram of the structure of an intelligent glare measurement system based on threshold increment according to an embodiment of the present invention;

[0047] Figure 2 A schematic diagram of a graphic brightness parameter extraction process according to an embodiment of the present invention;

[0048] Figure 3 Schematic diagram of the process of position information extraction and binocular vision correction according to an embodiment of the present invention;

[0049] Figure 4 Schematic diagram of system self-correction and feedback closed loop according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0051] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] The present embodiment proposes an intelligent glare measurement system based on threshold increment, comprising:

[0053] Data acquisition module, used for collecting road scene image data;

[0054] A data processing module, configured to extract brightness parameters and position information from the road scene image data;

[0055] A threshold increment calculation module is used to calculate the threshold increment based on the extracted brightness parameter and position information using a preset threshold increment calculation model;

[0056] The display and warning module is used to display the threshold increment and related data to the user, and automatically trigger the alarm mechanism when the measurement result exceeds the preset threshold.

[0057] Furthermore, the system further comprises:

[0058] The self-correction module is used to monitor the deviation between the measured data and the preset standard. When the deviation exceeds the standard, the feedback control algorithm is automatically called to dynamically adjust the image data acquisition parameters and image processing parameters, while optimizing the threshold increment calculation model to form a closed-loop control.

[0059] Specifically, to overcome the shortcomings of the existing system's hardware interface and unstable data transmission, this embodiment specifically designs a set of dedicated data acquisition and signal conditioning circuits based on the original system framework. This is the data acquisition module, which primarily performs image preprocessing, region segmentation, brightness and position information extraction, and conversion of images captured by industrial cameras into the required digital signals. This circuit module includes a preamplifier, a filter, an analog-to-digital converter (ADC), and a dedicated interface chip. This ensures that the analog signals output from the industrial camera can be transmitted to the embedded controller (i.e., the self-calibration and optimization feedback module) with high precision and low latency after high-speed conversion. The self-calibration and optimization feedback module is responsible for error monitoring and automatic parameter adjustment. The embedded controller (based on FPGA or ASIC technology) not only processes image data but also integrates real-time error monitoring and closed-loop feedback functions, capable of automatically adjusting image acquisition parameters (such as exposure time and gain) and image processing parameters according to environmental changes, thereby achieving high stability and accuracy for the entire system. The system mainly includes the following modules: data acquisition module, data processing module, threshold increment calculation module, display and warning module, and self-calibration module.

[0060] Furthermore, the data acquisition module realizes real-time acquisition of road scene image data through automatic exposure, automatic focus and parameter adjustment, and transmits the data to the data processing module;

[0061] The data acquisition module has a built-in preset parameter library and can automatically select acquisition mode and exposure parameters according to on-site conditions.

[0062] Specifically, in this embodiment, the data acquisition module is also called the image acquisition and transmission module:

[0063] Hardware composition: Use an industrial camera based on a CMOS sensor (such as MER-500-7UM), install it on a fixed bracket, and keep a vertical distance of about 1.5m from the road plane to ensure that the image acquisition is consistent with the actual observation position.

[0064] Functional description: This module realizes real-time acquisition of on-site images through automatic exposure, autofocus and parameter adjustment; it transmits the original image data to the computer control platform at high speed using data cables or wireless networks.

[0065] Additional functions: To cope with different lighting environments, the system has a built-in preset parameter library that can automatically select the appropriate acquisition mode and exposure parameters according to on-site conditions.

[0066] Furthermore, the data processing module includes:

[0067] Image preprocessing unit: used for preprocessing the road scene image data;

[0068] The brightness parameter extraction unit is used to combine the optical characteristics and imaging principles of the camera to establish an improved brightness calibration method. The pixel values of each area in the preprocessed image data are mapped to actual brightness values. The multi-point regression or nonlinear fitting algorithm is used to calculate the average brightness Lav of the road surface and the local brightness data of each light source area.

[0069] The position information extraction unit uses a binocular vision algorithm or a pre-calibrated monocular camera to automatically identify the position of lamps and other key light sources by performing image registration and feature point matching on pre-processed image data, extracting their position coordinates and area information. The extracted coordinates and area information serve as the basis for calculating the threshold increment.

[0070] Furthermore, the image preprocessing unit performs noise suppression, contrast enhancement, distortion correction and region segmentation processing on the road scene image data. During the preprocessing process, a filtering algorithm is used to eliminate random noise, and image enhancement technology is used to highlight the target area.

[0071] Specifically, in this embodiment, the data processing module is also called a digital image processing and preprocessing module:

[0072] Image preprocessing: noise suppression, contrast enhancement, distortion correction, and region segmentation are performed on the collected original images. Advanced filtering algorithms (such as median filtering and bilateral filtering) are used to eliminate random noise, and image enhancement techniques are used to highlight the target area.

[0073] Noise suppression: Median filtering and bilateral filtering algorithms are used to denoise the collected images to remove environmental noise and random interference.

[0074] Geometric correction: Utilizes pre-calibrated camera internal and external parameters to correct image distortion, ensuring that the image truly reflects the scene geometry during subsequent processing.

[0075] Image enhancement and segmentation: Adaptive histogram equalization algorithm is used to enhance image contrast, and then edge detection (such as Canny algorithm) and region segmentation technology are used to separate the target area (such as lamps and road surface) from the background, providing a clear area for brightness and position information extraction.

[0076] Brightness parameter extraction: Combining the optical characteristics and imaging principles of the camera, an improved brightness calibration method is established to map the pixel values of each area in the image to the actual brightness value (unit cd / m 2 ). Use multi-point regression or nonlinear fitting algorithm to calculate the average road surface brightness Lav and the local brightness data of each light source area.

[0077] Position Information Extraction: Utilizing binocular vision algorithms or combined with pre-calibrated monocular cameras, image registration and feature point matching are used to automatically identify the positions of lamps and other key light sources, extracting their 2D (or 3D) position coordinates and area information. This module fully considers factors such as road geometry and camera installation angle to ensure the accuracy and repeatability of position data.

[0078] Furthermore, the threshold increment calculation module includes:

[0079] A first calculation unit is used to calculate the equivalent light curtain brightness Lv for each light source using a pre-built mathematical model, and to cumulatively calculate the total light curtain brightness Lv,total;

[0080] The mathematical model pre-built here is the calculation model of the equivalent light curtain brightness, and the formula is as follows:

[0081]

[0082] The sum of k glare sources in the field of view is the total light curtain brightness L v , as shown in formula (3):

[0083]

[0084] Where: L v,k The equivalent light curtain brightness threshold increment value generated by the kth glare source, in cd / m 2 , E k The illuminance generated by the kth luminaire at the center of the observer's eye is in units of lx, θ k L is the angle between the line from the kth glare source to the observer's eyes and the line of sight, in degrees. vThe sum of k glare sources in the field of view is the total light curtain brightness, the unit is cd / m 2 .

[0085] The second calculation unit is configured to calculate the threshold increment by using a preset threshold increment calculation model.

[0086] Specifically, in this embodiment, the calculation model is constructed based on the Threshold Increment (TI) calculation formulas in international standard reports such as CIE31-1976 and CIE132-1999. This mathematical model comprehensively considers the light curtain brightness Lv, the total light curtain brightness Lv,total, and the average road surface brightness Lav. The model uses an integral accumulation method to superimpose the effects of various light sources on site to determine the total threshold increment.

[0087] Real-time online optimization: To address environmental changes and instrument drift that may occur in actual measurements, the module has a built-in closed-loop feedback algorithm that collects calculation errors in real time and automatically adjusts mapping model parameters based on a preset correction function to ensure that the threshold incremental error is controlled within 1%.

[0088] Data fusion: Using the brightness and position information extracted from the image processing module, each light source is independently calculated and then weighted fused to ensure that the overall glare measurement results fully reflect the actual situation on site.

[0089] The expression of the threshold increment calculation model is:

[0090] TI = f(Lv,total,Lav) (4)

[0091] Among them, TI is the threshold increment value, f is the mapping function, and the parameters are calculated using the following two formulas:

[0092]

[0093] Specifically, in this embodiment, the display and warning module:

[0094] Real-time display: The threshold increment value and related optical parameters are presented to the on-site operator in real time through the host computer software or embedded display screen. The on-site image and key measurement data are also displayed, making it easier for the operator to monitor the on-site conditions.

[0095] Data storage and transmission: The system stores all collected images and calculated data in local storage, and also supports transmission to a remote server or monitoring center through a network interface to facilitate subsequent data analysis and test report generation.

[0096] Alarm function: When the measurement result exceeds the preset threshold, the system can automatically trigger the alarm mechanism to prompt the operator to take appropriate measures to ensure road lighting safety.

[0097] Specifically, in this embodiment, the self-correction module is also called a system self-correction and optimization feedback module:

[0098] Error monitoring: The system regularly or in real time monitors the deviation between the measured data and the preset standard, and determines whether there is system parameter drift or environmental influence through statistical analysis and data comparison.

[0099] Automatic correction: When an error exceeding the standard is detected, the self-correction module automatically calls the feedback control algorithm to dynamically adjust the image acquisition parameters (such as exposure time and gain) and image processing parameters, while optimizing the threshold increment calculation model to form a closed-loop control.

[0100] Historical data records: All parameter adjustment records and system operation logs generated during the calibration process are archived to provide data support for subsequent system maintenance and technical improvements.

[0101] This module is implemented not only by software algorithms, but also by a dedicated hardware control chip for real-time error monitoring and feedback control. This dedicated control chip collects sensor data and, in conjunction with a pre-set correction circuit, automatically adjusts key parameters of the acquisition circuit, forming a closed-loop feedback system that integrates hardware and software.

[0102] The workflow of the entire system can be described as follows:

[0103] (1) Initialization phase:

[0104] After the system is powered on, it automatically loads the preset parameter library and performs initial self-calibration to determine the initial exposure, focus, and image processing parameters.

[0105] (2) Image acquisition and transmission:

[0106] Industrial cameras collect images in real time on site and transmit the data to the control platform.

[0107] (3) Image preprocessing and feature extraction:

[0108] The collected images are subjected to denoising, segmentation and enhancement processing to extract brightness information and position information respectively.

[0109] The brightness extraction module calculates the average brightness of the area Lav; the position extraction module extracts the lamp coordinates and area data through the binocular vision algorithm.

[0110] (4) Threshold increment calculation and optimization:

[0111] Using the pre-built mathematical model, the equivalent light curtain brightness Lv is calculated for each light source, and the total light curtain brightness Lv,total is calculated cumulatively;

[0112] Calculate the threshold increment according to the formula TI = f(Lv, total, Lav) and compare the result with the standard data;

[0113] If a deviation is detected, the self-correction feedback module is activated, the parameters are adjusted and recalculated until the preset accuracy is achieved.

[0114] (5) Result display and storage:

[0115] The measurement results are displayed on the operation terminal in real time and stored in the database;

[0116] Upload to the monitoring center via the network interface to facilitate subsequent data analysis and report generation.

[0117] (6) Closed-loop self-correction:

[0118] The system continuously monitors operating status and measurement errors, and automatically adjusts relevant parameters to ensure long-term stable operation.

[0119] In this embodiment, automated image acquisition and processing is implemented: industrial cameras are used to capture on-site images in real time, combined with advanced image preprocessing algorithms to achieve fully automated data acquisition without the need for manual deployment.

[0120] Accurate extraction of brightness and position information: Multiple filtering, region segmentation, and binocular vision technology are used to ensure high-precision extraction of image brightness and position parameters, providing a reliable basis for threshold increment calculation.

[0121] Real-time online optimization calculation model: Build a threshold increment calculation model based on international standards, and combine it with a closed-loop feedback mechanism to optimize the model in real time, so that the measurement accuracy is stabilized within 1%.

[0122] Self-correction feedback mechanism: Built-in error monitoring and self-correction module automatically adjusts system parameters through real-time data comparison to improve the system's adaptability and stability in different environments.

[0123] This embodiment uses industrial cameras to capture digital images and combines advanced image preprocessing, brightness and position information extraction, threshold increment calculation, and closed-loop self-correction technology to achieve automated, real-time, and high-precision measurement of road lighting glare. Its technical effects and data indicators are specifically reflected in the following aspects:

[0124] Automation and real-time performance:

[0125] (1) Fully automatic data collection:

[0126] Traditional methods require manual on-site deployment and instrument adjustment, and measurements typically take several hours. This system, however, uses industrial cameras to capture images in real time at a rate of up to 30 frames per second, shortening on-site data acquisition time by over 70%. Glare detection can be completed on large road areas within 5 minutes.

[0127] (2) Real-time data processing and feedback:

[0128] Using efficient image processing algorithms, the system completes image preprocessing, region segmentation, brightness and position information extraction, and threshold increment calculation within approximately 0.5 seconds after image acquisition, outputting measurement data in real time to ensure instant display of on-site data and meet the real-time needs of traffic safety monitoring.

[0129] High precision and data reliability:

[0130] (1) Brightness parameter extraction accuracy:

[0131] Experiments show that through an improved brightness calibration algorithm, the system can control the error in extracting the average brightness (Lav) of the road surface and lighting areas in the image within ±5%; under different lighting environments (such as daytime, nighttime, and rainy weather), the standard deviation of repeated measurement results is less than 0.2cd / m2, ensuring stable and reliable data.

[0132] (2) Position information extraction accuracy:

[0133] Using binocular vision and a pre-calibrated monocular camera method, the error in extracting the position parameters of the light source (lamp) is less than ±4%. For example, the error in the center position of the lamp is less than 0.1m, and the area measurement error is within ±5%, providing accurate geometric data support for the calculation of the threshold increment.

[0134] (3) Threshold increment calculation accuracy:

[0135] Based on the threshold increment (TI) calculation model constructed in the CIE technical report, multiple sets of experiments have verified that the threshold increment error calculated by the system can be controlled within ±1%, and the expanded uncertainty Urel is approximately 4.2% (k=2). This shows that in actual measurements, this system can reduce the error by 1-2 times compared to traditional methods, while maintaining high data repeatability.

[0136] Improved efficiency and reduced costs:

[0137] (1) Measurement efficiency is significantly improved:

[0138] Experimental data shows that under the same field conditions, traditional manual point measurement usually takes 4 hours to complete, while this system automatically collects and processes data in just 30 minutes to complete full-field data collection and preliminary processing, improving overall efficiency by more than 70%.

[0139] (2) Reduced equipment costs:

[0140] Traditional measurement requires the use of multiple professional instruments (illuminance meters, luminance meters, laser rangefinders, etc.), which have high equipment and maintenance costs. This system mainly relies on an industrial camera and a general computing platform, reducing the overall system cost by about 50% to 70%, making it easier to promote and apply on a large scale.

[0141] Improve site safety and serviceability:

[0142] (1) Reduce operational risks:

[0143] The use of non-contact automatic measurement method avoids surveyors from performing on-site operations in traffic-intensive areas, thereby reducing the risk of personnel being exposed to dangerous environments to almost zero.

[0144] (2) Wide range of applicable environments:

[0145] After field testing, the system can operate stably in different lighting environments such as highways, urban roads, tunnels and underground passages. The consistency of measurement data at night and in severe weather conditions remains above 95%, and its adaptability and stability are significantly improved compared to traditional methods.

[0146] Economic and social benefits:

[0147] (1) Energy conservation and emission reduction:

[0148] Real-time monitoring and automatic control can help management departments adjust the brightness and layout of street lights in a timely manner, reducing energy waste caused by excessive glare. It is estimated that the application of this system can reduce the energy consumption of road lighting systems by 10% to 15%.

[0149] (2) Improving traffic safety:

[0150] Accurate glare measurement data can provide a basis for optimizing road lighting design, reducing glare interference with drivers' vision and thus reducing nighttime traffic accidents. Actual test data shows that in the area where the system is used, the nighttime accident rate is expected to be reduced by more than 20%.

[0151] (3) Data standardization and intelligent management:

[0152] By establishing a road lighting measurement database, the high-precision data generated by the system can be used to build a standardized lighting evaluation system and intelligent management platform, providing scientific decision-making support for urban lighting planning and maintenance.

[0153] In summary, this embodiment achieves a closed-loop system from image acquisition to data processing, calculation, and real-time feedback through automated data acquisition, precise digital image processing, threshold increment calculation based on international standards, and a real-time self-correction feedback mechanism. This system not only significantly improves measurement accuracy and efficiency technically, but also demonstrates significant advantages in economic benefits, operational safety, and data management, providing an innovative and feasible solution for road lighting glare monitoring and traffic safety management.

[0154] like Figure 1 As shown in the figure, a hierarchical structure is used to illustrate the overall structure of the invention system and the data flow and connection relationship between each module. The figure uses rectangular boxes to represent the main modules, and arrows to indicate the direction of data flow. The relative position and hierarchy of each module are as follows:

[0155] Image acquisition module: Located in the upper left corner of the image, it uses an industrial camera to capture on-site images. The image contains descriptions such as "Industrial Camera (CMOS Sensor, MER-500-7UM)" and "Real-time On-site Image Acquisition."

[0156] Digital image processing module: located below (or to the right of) the image acquisition module, mainly used for image preprocessing, region segmentation, and brightness and position information extraction. Figure 1 In the module, indicate "denoising, enhancement, region segmentation, brightness calibration and position information extraction".

[0157] Threshold increment calculation module: It follows the digital image processing module, is located below or to the right of it, and indicates that the threshold increment (TI) is calculated using the CIE related formula, and also indicates "online parameter optimization and error control".

[0158] Display and feedback module: Located below the threshold increment calculation module, it is responsible for real-time display of measurement results and has data storage and remote transmission functions. The text description states "real-time data display, storage and alarm."

[0159] The Self-Correction and Optimization Feedback Module, located to the right or below the Display and Feedback Module, forms a closed-loop feedback loop. This module is subdivided into "Error Monitoring," "Feedback Control," and "Parameter Correction," represented in the diagram by multiple submodules or circular arrows, illustrating their closed-loop feedback relationship with the Display Module.

[0160] Figure 1 In the overall layout, each module is connected in sequence with arrow-shaped lines. The starting point of the arrow is the upstream data output, and the end point is the downstream data input, clearly showing the complete process of data from collection to processing, calculation, display and feedback.

[0161] Figure 2The processing flow of image brightness parameter extraction is described in detail, which mainly includes the following steps and modules. The layout is arranged from top to bottom, and the steps are connected by arrows:

[0162] Raw image acquisition: Located at the top of the figure, it describes the industrial camera collecting on-site image data and is marked with "raw image input".

[0163] Image preprocessing: Located below the original image acquisition module, it includes processing steps such as denoising, correction (such as distortion correction) and image enhancement, marked as "noise suppression, contrast enhancement".

[0164] Region Segmentation: Immediately below the image preprocessing module, it describes the segmentation of target areas such as lamps and roads from the background in the image, marked as "Target Region Segmentation".

[0165] Brightness Calibration: Located below the region segmentation module, this section indicates converting pixel values to actual brightness (cd / m2) using a pre-established brightness calibration method. It is labeled "Brightness Calibration, Pixel Value Mapping."

[0166] Brightness data regression calculation: Located at the bottom of the figure, it describes how to obtain parameters such as the regional average brightness Lav through regression or fitting algorithms, and is marked with "Brightness regression calculation, regional average brightness Lav".

[0167] The arrows between each step clearly indicate the processing flow from data acquisition to final brightness parameter extraction, ensuring that the logical relationship and data conversion process of each processing link are clear and accurate.

[0168] Figure 3 This demonstrates how to use binocular vision (or combined with a calibration plate method) to automatically extract and calibrate the position information of on-site light sources (such as lamps). The layout of the modules in the figure is as follows:

[0169] Image acquisition: Located at the top of the figure, it indicates the acquisition of binocular or monocular images and the combination with the preset calibration plate image.

[0170] Binocular Image Matching / Calibration Correction: Immediately below the image acquisition module, this section describes how to calibrate internal and external parameters using a binocular matching algorithm or a calibration plate to ensure image alignment. It is labeled "Internal and External Parameter Calibration, Binocular Matching."

[0171] Feature Point Detection: Located below the binocular matching module, this function detects the edges and feature points of lamps or light sources in the image and is labeled "Lamp Feature Extraction, Edge Detection."

[0172] Position information calculation: Immediately below feature point detection, use the matching results to calculate the 2D or 3D coordinates, area, and relative distance of the light source, marked "Position information acquisition, area and distance calculation."

[0173] Correction processing: It is set last after the position information calculation module, indicating that the calculation results are corrected and adjusted through the preset model to ensure measurement accuracy, and marked as "position correction, parameter optimization".

[0174] The modules are connected by arrows, and the process clearly shows the complete process from acquisition, matching, feature detection to position information calculation and correction.

[0175] Figure 4 The working principle of the system's self-correction and feedback closed loop is demonstrated, and a decision selection box is introduced to distinguish whether the system parameters need to be adjusted. The layout in the figure is as follows:

[0176] Measurement data output module: Located at the top of the diagram, it represents the threshold increment (TI) and related data currently output by the system.

[0177] Error monitoring module: located immediately below the measurement data output module, it describes how to compare actual data with preset standards and calculate the measurement error.

[0178] Decision selection box (diamond): Located below the error monitoring module, it displays the judgment "Is the error exceeding the standard?" Two paths are branched from this decision box: one marked "yes" points to the feedback control module; the other marked "no" points to the normal operation process (directly fed back to the data output).

[0179] Feedback control module: Located below the decision box in the "Yes" branch, it is responsible for generating feedback signals based on the error and triggering subsequent correction processes.

[0180] Correction update module: Located below the feedback control module, it automatically adjusts image acquisition and processing parameters according to feedback signals and performs system parameter correction.

[0181] Closed-loop feedback: The output of the correction update module is fed back to the measurement data output module through an arrow, forming a closed loop. In the "No" branch, the current data is directly retained and output.

[0182] The entire process uses arrows to indicate the connection and data flow of each module, and the decision selection box clearly distinguishes whether to trigger the self-correction function, ensuring that the system can achieve closed-loop feedback and dynamic adjustment.

[0183] This example describes in detail the construction and operation of an intelligent glare measurement system based on threshold increments, using a practical experimental environment. By integrating an industrial camera, image acquisition and processing software, and an embedded controller, this system achieves automated, real-time, and high-precision road lighting glare measurement. The following details the hardware architecture, software algorithms, data acquisition and processing, threshold increment calculation, and closed-loop self-calibration.

[0184] Hardware composition:

[0185] The hardware of this system mainly includes industrial cameras, brackets, data transmission modules and control computers. Its specific configuration is as follows:

[0186] Industrial cameras:

[0187] This embodiment uses the MER-500-7UM industrial camera produced by China Daheng, which is based on a progressive scan CMOS image sensor with a resolution of 2592×1944 and a pixel size of 2.2 μm×2.2 μm.

[0188] The camera supports a standard USB 2.0 data interface and features automatic exposure, autofocus, and adjustable gain, ensuring high-quality image data capture in varying lighting conditions. During installation, the camera is fixed to a bracket, perpendicular to the road surface, at a height of 1.5 meters to simulate the typical position of a driver's eyes.

[0189] Bracket and data transmission:

[0190] The bracket is made of high-strength alloy material to ensure stability and wind resistance in outdoor environments.

[0191] Data transmission uses USB2.0 or wireless transmission module to transmit the collected images to the control computer in real time, realizing data reception without delay.

[0192] Control computer:

[0193] The control computer is equipped with a high-performance processor and sufficient storage space to run image acquisition and processing software, data analysis algorithms and self-correction feedback systems.

[0194] Pre-installed dedicated host computer software to achieve real-time image acquisition, data preprocessing, parameter calculation and measurement result display, and has data storage and network transmission functions.

[0195] In addition to utilizing the MER-500-7UM industrial camera, this system also features a dedicated data acquisition and signal conditioning circuit. This circuit, consisting of a preamplifier, a low-pass filter, and a high-precision ADC, converts the analog camera output signal into a digital signal and performs anti-interference processing. This data acquisition circuit is integrated with an FPGA-based embedded controller, connected via a dedicated high-speed interface, enabling real-time, high-speed transmission of image data. Furthermore, the embedded controller incorporates a built-in closed-loop feedback control algorithm that monitors the quality of collected data in real time and automatically adjusts key parameters of the acquisition circuit to ensure stable and accurate data transmission even in complex field environments.

[0196] Software algorithm and data processing flow:

[0197] The software part of this system consists of an image acquisition module, a digital image processing module, a threshold increment calculation module, and a closed-loop self-correction feedback module. Its main workflow is divided into the following steps:

[0198] Image acquisition:

[0199] After the control software is started, the on-site images are collected in real time through the industrial camera.

[0200] The image acquisition software automatically adjusts the exposure and focus parameters and transmits the acquired raw images to the image processing module at a speed of 30 frames per second to ensure data continuity.

[0201] Image preprocessing:

[0202] Noise suppression: Median filtering and bilateral filtering algorithms are used to denoise the collected images to remove environmental noise and random interference.

[0203] Geometric correction: Utilizes pre-calibrated camera internal and external parameters to correct image distortion, ensuring that the image truly reflects the scene geometry during subsequent processing.

[0204] Image enhancement and segmentation: Adaptive histogram equalization algorithm is used to enhance image contrast, and then edge detection (such as Canny algorithm) and region segmentation technology are used to separate the target area (such as lamps and road surface) from the background, providing a clear area for brightness and position information extraction.

[0205] Brightness parameter extraction:

[0206] After region segmentation, the pixel values in the image are converted into actual brightness values (unit: cd / m2) through brightness calibration method.

[0207] The average brightness Lav of the target area is calculated using a regression algorithm (linear or nonlinear fitting). Experimental data show that the extraction error of this value can be controlled within ±5%.

[0208] Location information extraction

[0209] Using binocular vision or a monocular camera combined with a correction plate, the two-dimensional position, area and relative distance of key light sources such as lamps are extracted through image feature matching and geometric calculation.

[0210] Experiments show that the position information error of this method is typically less than ±4%, providing accurate spatial geometric data for threshold increment calculation.

[0211] Threshold increment calculation:

[0212] A mathematical model constructed according to international standards such as CIE31-1976 and CIE132-1999 uses the extracted average road surface brightness Lav, the equivalent light curtain brightness of each light source Lv, and the total light curtain brightness Lv,total to calculate the threshold increment using the formula TI = f(Lv,total,Lav).

[0213] The model uses an integral accumulation method to superimpose the effects of all effective light sources in the field of view. Experimental data show that the error of the final calculated TI value can be controlled within ±1%, and the expanded uncertainty is about 4.2% (k=2).

[0214] Closed-loop self-correction and feedback:

[0215] The system has a built-in error monitoring algorithm that compares real-time measurement data with preset standards and automatically generates a feedback signal when it detects a deviation exceeding the set range.

[0216] The feedback control module calls the self-correction algorithm based on the feedback signal, adjusts the image acquisition parameters (such as exposure time and gain) and image processing parameters (such as filter weight and segmentation threshold), and updates the threshold increment calculation model in real time to form a closed-loop feedback to ensure the long-term stable operation of the system.

[0217] The corrected data re-enters the data processing process, forming a cycle until the measured data is stable and meets the requirements.

[0218] In this system, the embedded controller not only performs image preprocessing and data analysis but also provides real-time error monitoring and closed-loop feedback control. Using a built-in dedicated circuit module, the controller collects data converted by the ADC in real time and transmits it to the host software for subsequent threshold increment calculation. Furthermore, if the system detects data deviations exceeding a preset range, the embedded controller generates a feedback signal to directly adjust hardware parameters (such as camera exposure, gain, and acquisition circuit filter coefficients), ensuring closed-loop self-correction for the entire system.

[0219] Example 1: Intelligent glare measurement under standard environment:

[0220] In an urban road environment, the system is deployed according to the solution of this embodiment.

[0221] Experimental setup: An industrial camera was installed 1.5 meters from the road to collect image data including the road surface, street lights, and oncoming vehicles. Automatic exposure mode was used for image acquisition, and the frame rate was set to 30 fps.

[0222] Image processing: In image preprocessing, median filter and bilateral filter are used to suppress image noise, and the lamp area and road area are extracted after region segmentation.

[0223] Brightness calibration: The average road surface brightness Lav was calculated using a regression fitting method. The actual Lav value measured in the test was 1.2 cd / m2, with an error of ±0.06 cd / m2.

[0224] Position information: The binocular vision algorithm is used to extract the position of the lamp, and the error of the lamp center position is less than 0.1m.

[0225] Threshold increment calculation: Using integral accumulation calculation, the total light curtain brightness Lv,total is calculated to be 3.5cd / m2, corresponding to the threshold increment TI value of 1.8%, and the measurement error is less than ±1%.

[0226] Self-calibration feedback: During continuous operation, if the TI value is detected to deviate from the standard value by more than 1.0%, the system automatically adjusts the camera exposure and image processing parameters. After self-calibration, the TI value is stabilized at 1.8% ± 0.02%.

[0227] Example 2: Self-correction application in complex light environment:

[0228] In the highway tunnel environment, the lighting conditions are complex, with obvious local highlights and low-light areas.

[0229] Experimental setup: The camera is fixed and uses built-in environmental sensors to obtain parameters such as temperature and humidity to assist in calibration.

[0230] Image acquisition and preprocessing: During the image acquisition process, the system automatically detects changes in ambient light and dynamically adjusts filter parameters to achieve stable image quality.

[0231] Brightness and position information extraction: After multiple tests, the system controls the brightness data extraction error within ±5%, and the position information error within ±4%.

[0232] Threshold increment calculation and feedback: After the initial calculation, if the TI value deviates by ±1.5% due to ambient light fluctuations, the self-correction module is triggered. The system automatically adjusts the TI value to 1.5% ±0.01% to ensure data accuracy.

[0233] Long-term stable operation: In a 6-hour continuous operation test, the system effectively eliminated measurement errors caused by environmental changes through closed-loop self-correction, ensuring data stability of more than 95%.

[0234] Implementation effect:

[0235] As can be seen from the above embodiments, the system can accurately and in real time measure glare parameters in different light environments. Its main technical effects are reflected in:

[0236] The error in image brightness parameter extraction is less than ±5%, the error in position information is less than ±4%, and the error in threshold increment calculation is controlled within ±1%;

[0237] Measurement efficiency has increased by more than 70%, and on-site measurement time has been shortened from the traditional 3 to 4 hours to 20 to 30 minutes;

[0238] System automation and closed-loop self-correction significantly reduce on-site operation risks and improve data continuity and reliability;

[0239] In terms of economic cost, the overall system uses general industrial cameras and general computing platforms. The equipment investment and maintenance costs are 50% to 70% lower than traditional instruments, and it has good prospects for promotion and application.

[0240] This embodiment specifically relates to a threshold increment-based intelligent glare measurement system—a closed-loop self-correction solution integrating hardware and software. This system not only utilizes industrial cameras and advanced image processing algorithms to automatically detect and eliminate glare on site, but also utilizes dedicated data acquisition and signal conditioning circuits, an embedded controller, and closed-loop feedback control circuits to implement real-time signal conditioning and self-correction at the hardware level, significantly improving the stability, real-time performance, and accuracy of measurement data.

[0241] This system integrates dedicated hardware circuits (including pre-amplifiers, low-pass filters, and high-precision analog-to-digital converters) with an embedded controller (based on FPGA or ASIC technology) to form a closed-loop control system with coordinated hardware and software. The system can acquire and process road scene image information in real time, automatically extract image brightness and position information, and perform data calculations based on a threshold increment calculation model established according to international standards. Furthermore, a built-in dedicated hardware feedback circuit enables real-time error monitoring and automatic parameter adjustment, ensuring that the entire measurement process maintains high accuracy and stability even in complex environments.

[0242] Therefore, this embodiment is suitable for automatic, real-time, and high-precision measurement of glare in various complex lighting environments, such as traffic roads, tunnels, and underground passages. Its core innovation lies in the deep integration of software and hardware, dedicated data acquisition and signal conditioning technology, and closed-loop self-correction feedback control scheme.

[0243] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. An intelligent glare measurement system based on threshold increment, characterized in that: include: Data acquisition module, used for collecting road scene image data; A data processing module, configured to extract brightness parameters and position information from the road scene image data; A threshold increment calculation module is used to calculate a threshold increment value based on the extracted brightness parameter and position information using a preset threshold increment calculation model; The display and warning module is used to display the threshold increment value and related data to the user, and automatically trigger the alarm mechanism when the threshold increment value exceeds the preset threshold.

2. The intelligent glare measurement system based on threshold increment according to claim 1, characterized in that: The system further comprises: The self-correction module is used to monitor the deviation between the measured data and the preset standard. When the deviation exceeds the standard, the feedback control algorithm is automatically called to dynamically adjust the image data acquisition parameters and image processing parameters, while optimizing the threshold increment calculation model to form a closed-loop control.

3. The intelligent glare measurement system based on threshold increment according to claim 1, characterized in that: The data acquisition module realizes real-time acquisition of road scene image data through automatic exposure, automatic focus and parameter adjustment, and transmits the data to the data processing module; The data acquisition module has a built-in preset parameter library and can automatically select acquisition mode and exposure parameters according to on-site conditions.

4. The intelligent glare measurement system based on threshold increment according to claim 1, characterized in that: The data processing module includes: Image preprocessing unit: used for preprocessing the road scene image data; The brightness parameter extraction unit is used to combine the optical characteristics and imaging principles of the camera to establish an improved brightness calibration method. The pixel values of each area in the preprocessed image data are mapped to actual brightness values. The multi-point regression or nonlinear fitting algorithm is used to calculate the average brightness Lav of the road surface and the local brightness data of each light source area. The position information extraction unit is used to automatically identify the position of lamps and other key light sources and extract position coordinates and area information by performing image registration and feature point matching on the pre-processed image data using a binocular vision algorithm or in combination with a pre-calibrated monocular camera method.

5. The intelligent glare measurement system based on threshold increment according to claim 4, characterized in that: The image preprocessing unit performs noise suppression, contrast enhancement, distortion correction and region segmentation processing on the road scene image data. During the preprocessing process, a filtering algorithm is used to eliminate random noise, and an image enhancement technology is used to highlight the target area.

6. The intelligent glare measurement system based on threshold increment according to claim 1, characterized in that: The threshold increment calculation module includes: A first calculation unit is used to calculate the equivalent light curtain brightness Lv for each light source using a pre-built mathematical model, and to cumulatively calculate the total light curtain brightness Lv,total; The second calculation unit is configured to calculate a threshold increment value by using a preset threshold increment calculation model.

7. The intelligent glare measurement system based on threshold increment according to claim 1, characterized in that: The expression of the threshold increment calculation model is: TI=f(Lv,total,Lav) Where TI is the threshold increment value and f is the mapping function.

8. The intelligent glare measurement system based on threshold increment according to claim 2, characterized in that: The self-correction module includes: Error monitoring unit, used to monitor the deviation between the measured data and the preset standard, and determine whether there is system parameter drift or environmental influence through statistical analysis and data comparison; Automatic correction unit, which is used to automatically call the feedback control algorithm when the error exceeds the limit, dynamically adjust the image data acquisition parameters and image data processing parameters, and optimize the threshold increment calculation model to form a closed-loop control; The historical data recording unit is used to archive the parameter adjustment records and system operation logs generated during the calibration process.

Images