A smart city street lamp lighting adjustment method and system

Abstract

The application relates to the technical field of street lamp lighting, and discloses a smart city street lamp lighting adjusting method and system. The method obtains a road image through an image sensor on a street lamp, extracts a vehicle feature image such as a vehicle lamp and a glass distribution, combines multi-dimensional matching value calculation to determine a real vehicle and determine a driving direction; then, a feature speed is calculated through feature image displacement and a time stamp, if the speed exceeds a set speed, a speed contrast value is calculated, and the brightness of the street lamp in a set distance in front of the vehicle is controlled to change at intervals. When adjusting, the brightness is inversely related to the distance between the street lamp and the vehicle, and the brightness fluctuation amplitude is negatively related to the speed contrast value. The application can dynamically adapt to the driving state of the vehicle, avoid invalid lighting, ensure driving safety and visual comfort, and reduce energy waste.

Description

Technical Field

[0001] This application relates to the technical field of street lighting, and in particular to a smart city street lighting adjustment method and system. Background Technology

[0002] Urban street lighting is a core component of urban infrastructure, playing a crucial role in ensuring nighttime traffic safety and enhancing security in public areas. Its technological system has evolved through multiple generations, gradually upgrading from traditional incandescent and high-pressure sodium lamps to high-efficiency LED-based light source systems, and forming a supporting infrastructure including power supply modules, control units, and wiring networks.

[0003] With the advancement of smart city construction, urban street lighting technology is developing towards energy conservation. Intelligence is reflected in the introduction of control center technology, which determines different lighting times based on varying sunlight conditions, enabling remote monitoring and centralized management of street light status. Energy conservation focuses on optimizing energy consumption by combining environmental data to respond to urban construction needs.

[0004] Currently, the vast majority of urban street lighting systems still employ a constant lighting mode, meaning they operate by pre-setting fixed power or brightness parameters. While these systems offer simple adjustments based on time or ambient light intensity, they lack sophisticated methods to adapt to real-time changes in road conditions. In real-world scenarios, road conditions vary significantly, making constant lighting modes ill-suited to the refined management needs of smart cities. Summary of the Invention

[0005] To improve the precision of urban street lighting adjustment to suit road conditions, this application provides a smart urban street lighting adjustment method and system.

[0006] Firstly, this application provides a method for adjusting the lighting of smart city streetlights, employing the following technical solution:

[0007] A method for adjusting street lighting in a smart city includes the following steps:

[0008] Road images are acquired using image sensors mounted on streetlights, and a first vehicle feature image and a second vehicle feature image are extracted from the road images.

[0009] A first image matching value is calculated based on the first vehicle feature image and a preset first feature template; a second image matching value is calculated based on the second vehicle feature image and a preset second feature template; and a feature position matching value is calculated based on the first vehicle feature image and the second vehicle feature image.

[0010] The vehicle matching value is calculated based on the first image matching value, the second image matching value, and the feature position matching value. If the vehicle matching value is greater than the preset matching value, the vehicle direction is calculated based on the first vehicle feature image and the second vehicle image, and the first continuous timestamp of the first vehicle feature image and the second continuous timestamp of the second vehicle feature image are obtained.

[0011] A first speed is calculated based on the displacement of the first vehicle feature image on the road image and a first consecutive timestamp; a second speed is calculated based on the displacement of the second vehicle feature image on the road image and a second consecutive timestamp; a characteristic speed is calculated based on the first and second speeds; if the characteristic speed is greater than a preset speed, a speed comparison value is calculated based on the characteristic speed and the preset speed, and the brightness interval of the streetlights within a preset distance in front of the current vehicle is controlled.

[0012] The brightness of the streetlights is adjusted inversely based on the distance between the streetlights and the vehicles in front of them; the farther the streetlights are, the lower the brightness; the closer the streetlights are, the higher the brightness.

[0013] The brightness fluctuation amplitude is adjusted based on the negative correlation between the speed reference value and the brightness interval. The larger the speed reference value, the smaller the brightness fluctuation amplitude, and vice versa.

[0014] By adopting the above technical solution, vehicles can be accurately identified. By extracting dual features of headlight distribution and glass distribution, and combining image matching values ​​and feature position matching values, vehicle matching values ​​are calculated to avoid misjudgment based on a single feature and ensure that adjustments are triggered only for real vehicles. The lighting can be dynamically adapted to vehicle status. The feature speed is obtained by integrating dual speed verification and adjusting the brightness fluctuation range according to the speed comparison value. At high speeds, fluctuations are reduced to ensure stable vision, while fluctuations are increased at low speeds to facilitate observation of details. At the same time, the brightness is adjusted according to the inverse correlation between the street light and the vehicle distance to form a gradient lighting zone, taking into account both driving safety and environmental comfort. It can also achieve on-demand lighting, maintaining a basic low brightness when there are no vehicles and adjusting the street lights in front only when there are vehicles, which greatly reduces energy waste and maintenance costs.

[0015] Optionally, the method further includes the following steps:

[0016] Calculate the vehicle feature size parameters based on the first vehicle feature image and the second vehicle feature image;

[0017] The vehicle size is matched based on the vehicle characteristic size parameters;

[0018] The size reference value is calculated based on the vehicle size and the preset size.

[0019] The setting distance is adjusted according to the size reference value. The larger the size reference value, the longer the setting distance, and the smaller the size reference value, the shorter the setting distance.

[0020] By adopting the above technical solutions, large vehicles have a larger reference value for size, which extends the set distance to ensure lighting and improve safety; small vehicles have a smaller reference value, which shortens the distance, ensuring lighting while reducing energy consumption, thus achieving precise and on-demand lighting.

[0021] Optionally, the method further includes the following steps:

[0022] Extract shadow and road surface images from road images;

[0023] Calculate the shadow pixel values ​​of the shadow image and the road surface pixel values ​​of the road surface image;

[0024] The shadow reference value is calculated based on the shadow pixel value and the road surface pixel value according to a preset algorithm. The shadow reference value = ((road surface pixel value - shadow pixel value) / pixel value dynamic range) × reference value scaling factor;

[0025] The base brightness value of the street light is adjusted according to the positive correlation with the shadow reference value. The larger the shadow reference value, the higher the base brightness value of the street light; the smaller the shadow reference value, the lower the base brightness value of the street light.

[0026] By adopting the above technical solutions, a large shadow reference value increases the base value of brightness to compensate for shadow lighting; a small value reduces the base value to ensure lighting while saving energy, and matches the actual lighting to reduce visual discomfort.

[0027] Optionally, the method further includes the following steps:

[0028] Calculate the rate of change of the speed reference value within a set period;

[0029] The minimum adjustment unit for brightness fluctuation amplitude is adjusted based on the negative correlation between the rate of change and the rate of change. The faster the rate of change, the larger the minimum unit, and the slower the rate of change, the smaller the minimum unit.

[0030] By adopting the above technical solutions, when the brightness contrast value changes rapidly, the smallest unit of brightness fluctuation adjustment is increased to quickly respond and prevent lighting lag; when the change is slow, the unit is decreased to achieve smooth adjustment, avoiding frequent fine-tuning that causes visual fatigue and ensuring driving visibility.

[0031] Optionally, the method further includes the following steps:

[0032] Calculate the trend value of the speed reference value, where the trend value is the rate or amount of change of the speed reference value over a continuous time interval, and adjust the color temperature of the street light according to the trend value.

[0033] If the trend value is greater than the value within the preset trend value range, the color temperature of the street light will be the preset low color temperature value.

[0034] If the trend value is within the preset trend value range, the color temperature of the street light is the preset medium color temperature value.

[0035] If the trend value is less than the value within the preset trend value range, then the color temperature of the street light is the preset high color temperature value.

[0036] By adopting the above technical solutions, a medium color temperature is used when the speed is stable, which is soft and comfortable and reduces fatigue; a low color temperature is used when accelerating, with warm light enhancing the sense of three-dimensionality and helping to capture details; and a high color temperature is used when decelerating, with cold light having strong penetrating power to improve clarity and facilitate observation of the surroundings.

[0037] Optionally, the method further includes the following steps:

[0038] Calculate the trend value of the speed reference value, where the trend value is the rate or amount of change of the speed reference value over a continuous time interval;

[0039] The calculated value of the changing trend is determined based on the changing trend value and the preset reference trend value.

[0040] The color temperature of the streetlights is adjusted based on the inverse correlation of the calculated value of the trend. The smaller the calculated value of the trend, the brighter the color temperature of the streetlight; the larger the calculated value of the trend, the darker the color temperature of the streetlight.

[0041] By adopting the above technical solution, the color temperature is adjusted by dynamically calculating the trend value. The color temperature is brighter when the trend value is small and darker when the trend value is large, which accurately adapts to changes in vehicle speed and improves driving comfort and safety.

[0042] Optionally, the method further includes the following steps:

[0043] Acquire laser-sensing images of the light-shaped beams scanned by the vehicle's LiDAR;

[0044] The size of the coverage area is calculated based on the laser-sensed image;

[0045] The number of streetlights in each interval is adjusted according to the positive correlation between the coverage area and the brightness interval. The larger the coverage area, the more streetlights are in each interval; the smaller the coverage area, the fewer streetlights are in each interval.

[0046] By adopting the above technical solutions, when the coverage area is large, increasing the number of streetlights per interval can ensure large-scale lighting coverage and avoid lighting gaps; when the coverage area is small, reducing the number of streetlights per interval can avoid excessive lighting waste.

[0047] Optionally, the method further includes the following steps:

[0048] Obtain the scanning light resolution of the laser-sensing image;

[0049] The brightness increment is adjusted in a positive correlation with the scanning light resolution. The higher the scanning light resolution, the greater the brightness increment; the lower the scanning light resolution, the smaller the brightness increment.

[0050] By adopting the above technical solutions, when the resolution is high (more detailed capture of surrounding details, such as complex road conditions), increasing the brightness increment can enhance the lighting of key areas and improve the recognition of details; when the resolution is low (simple environment), reducing the brightness increment can avoid over-illumination and save energy while ensuring basic visibility.

[0051] Optionally, the method further includes the following steps:

[0052] Acquire the scanning light frequency of the laser-sensing image;

[0053] The brightness reduction for dimming is adjusted according to the positive correlation between the scanning light frequency and the brightness. The higher the scanning light frequency, the greater the brightness reduction for dimming; the lower the scanning light frequency, the smaller the brightness reduction for dimming.

[0054] By adopting the above technical solutions, when the frequency is high, increasing the brightness reduction can quickly reduce the brightness of areas that do not require lighting, thus reducing energy waste; when the frequency is low, decreasing the brightness reduction can achieve a smooth transition in brightness, avoiding frequent changes in brightness that could affect visual comfort.

[0055] Secondly, this application provides a smart city street light adjustment system, which adopts the following technical solution:

[0056] A smart city street lighting control system includes a processor, wherein the processor performs the steps of the smart city street lighting control method as described in any one of the preceding claims.

[0057] In summary, this application includes at least one of the following beneficial technical effects: By dynamically adjusting street lighting in multiple dimensions, it achieves accurate vehicle recognition, scene-specific lighting adaptation, and efficient energy utilization; by combining dual-feature matching and multi-parameter verification, it ensures that adjustments are triggered only for real vehicles, avoiding misjudgments; based on vehicle size, speed, illumination, and laser sensing data, it dynamically adapts to different vehicle types, road conditions, and driving states from multiple aspects such as brightness, color temperature, and adjustment range, taking into account both driving safety and driving comfort; by adjusting the lighting range and intensity as needed, it significantly reduces ineffective lighting, thereby reducing energy consumption and light pollution. Attached Figure Description

[0058] Figure 1 This is a step-by-step diagram of a smart city street light adjustment method.

[0059] Figure 2 This is a step-by-step diagram showing how to dynamically adjust the lighting coverage based on vehicle dimensions. Detailed Implementation

[0060] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0061] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0062] This application discloses a method for adjusting street lighting in a smart city, referring to... Figure 1 It includes the following steps:

[0063] High-definition image sensors, such as industrial-grade CMOS sensors, are deployed on streetlights to acquire real-time road images covering their monitoring range. These sensors possess high dynamic range and low light sensitivity, enabling them to clearly capture details of vehicles, pedestrians, and the environment under complex lighting conditions, including bright sunlight in sunny days and low light at night, providing high-quality image data for subsequent feature extraction. In the image preprocessing stage, the system eliminates image noise, such as road stains and glare interference, through edge detection and grayscale conversion algorithms. From the processed road images, two types of vehicle feature images are extracted: the first is a primary vehicle feature image, prioritizing vehicle headlight distribution features, such as the shape, brightness, and spacing of headlights and taillights. Headlights have high recognizability in both day and night environments and can serve as markers for vehicle location. The second is a secondary vehicle feature image, prioritizing vehicle glass distribution features, such as the outline and transmittance of the windshield and side windows. The significant pixel differences between the glass and the vehicle body can help verify vehicle integrity. This combined extraction of two types of features overcomes the limitations of single features and reduces misjudgments.

[0064] To avoid misidentifying non-vehicle targets, such as roadside trees and temporary obstacles, as vehicles, the system uses a triple matching value calculation, with the following determination logic:

[0065] The extracted first vehicle feature image is compared with a preset first feature template to calculate similarity, outputting a first image matching value ranging from 0 to 100. A higher value indicates a higher degree of matching of the vehicle headlight features. The first feature template is a standard template library built upon massive vehicle headlight data, covering headlight feature parameters for mainstream models such as sedans, trucks, and SUVs. Similarly, the second vehicle feature image is compared with a preset second feature template, outputting a second image matching value. This dual single-feature matching provides an initial screening of vehicle targets at the local feature level. The second feature template is a template library built upon the glass distribution patterns of different vehicle models.

[0066] Based on the physical structure of vehicles, such as headlights typically being located at the front / rear of the vehicle body and glass at the middle, and the two having a fixed relative positional relationship, the pixel spacing and angular deviation of the first vehicle feature image and the second vehicle feature image in the road image are calculated, and the positional matching value between the features is output. If the positional relationship between the two conforms to the structural logic of the real vehicle, such as the distance between the headlights and the glass being within a preset reasonable range, the positional matching value is high, otherwise it is low. This step eliminates the possibility of misjudging non-vehicle targets from the overall structural level. For example, if a distant street light spot is mistakenly combined with a roadside glass billboard, their positional relationship will inevitably deviate from the vehicle's structural rules.

[0067] The system uses a weighted algorithm to fuse the first image matching value, the second image matching value, and the feature position matching value to calculate the final vehicle matching value. The weighted algorithm dynamically adjusts the weights based on day and night conditions; for example, it gives higher weight to vehicle headlight features at night and higher weight to glass features during the day. The system determines the current target as a real vehicle only when the vehicle matching value is greater than a preset matching value. This preset matching value is a threshold calibrated through extensive real-world testing, typically set at 80, to ensure high accuracy even in complex environments. If the matching value is below the threshold, the target is determined to be non-vehicle, and subsequent lighting adjustments are not triggered, effectively avoiding ineffective adjustments and reducing energy waste and redundant system computation.

[0068] After confirming the actual vehicle target, the system further obtains the vehicle's driving direction and speed to provide dynamic basis for lighting adjustment:

[0069] Based on the relative positional relationship between the first and second vehicle feature images, if the headlights are in front of the glass, the vehicle's direction of travel can be determined to be away from the image sensor; if the headlights are behind the glass, it can be determined to be close to the image sensor. Combining the coordinate system of the road image, a two-dimensional coordinate system is established with the street light position as the origin to accurately output the vehicle's direction of travel, such as traveling from east to west or from south to north. This ensures that subsequent adjustments are made only to the street lights in front of the vehicle, avoiding ineffective adjustments to the street lights behind or to the side of the vehicle.

[0070] Using time-series data from image sensors, the first consecutive timestamp of the first vehicle feature image and the second consecutive timestamp of the second vehicle feature image are obtained. The first consecutive timestamps are such as t1 and t2, two consecutive sampling times. The displacement of the two types of feature images in the road image coordinate system is calculated respectively. For example, at time t1, the headlight is at coordinates (x1, y1), and at time t2, it is at (x2, y2). The displacement is the pixel distance between the two points, which is then converted to the actual road surface distance through pixel ratio. According to the formula: speed = displacement / time difference, the first speed based on the headlight displacement and the second speed based on the glass displacement are calculated respectively. Finally, the average of the two or the effective value after removing outliers is taken as the feature speed. Dual speed verification can avoid the error of single feature displacement calculation, such as miscalculation of headlight displacement caused by temporary obstruction due to bumps, and ensure the accuracy of speed data.

[0071] If the characteristic speed is greater than the preset speed, such as 5 km / h, excluding scenarios where the vehicle is stationary or traveling at extremely low speeds (which do not require dynamic lighting adjustment), then the speed reference value is calculated using the formula: Speed ​​reference value = (characteristic speed - set speed) / preset maximum speed × 100. For example, if the maximum speed is set to 120 km / h, a higher speed reference value indicates a faster vehicle speed. If the characteristic speed is lower than the set speed, then the basic brightness of the streetlights is maintained to avoid excessive adjustment in low-speed scenarios that could affect the driver's vision.

[0072] After capturing the vehicle's status, the system dynamically adjusts the streetlights within a set distance in front of the vehicle. This set distance can be preset according to the road level, such as 500 meters for urban main roads and 200 meters for community side roads. The system adjusts the distance dynamically based on two dimensions: brightness gradient and fluctuation amplitude.

[0073] By using the ranging function of the image sensor or combining it with preset spacing data between streetlights, the distance between each target streetlight and the current vehicle is calculated, and the brightness is adjusted according to an inverse correlation rule: the farther the streetlight is from the vehicle, the lower the brightness; for example, the brightness of a streetlight at 500 meters is set to 30% of the base brightness, providing only ambient outline illumination. The closer the streetlight is to the vehicle, the higher the brightness; for example, the brightness of a streetlight at 100 meters is set to 80% of the base brightness, providing clear road illumination. Through this gradient lighting mode of brighter near and darker far, glare caused by direct glare from the front of the vehicle is avoided, as full brightness from all streetlights can easily cause driver fatigue. At the same time, continuous lighting is ensured along the vehicle's driving path, forming a dynamic lighting corridor that adapts to the driver's line of sight.

[0074] Considering the different visual needs of drivers at different driving speeds, stable lighting is needed at high speeds to avoid visual interference, while sensitive lighting is needed at low speeds to capture road details. The system adjusts the brightness fluctuation range of brightness interval changes according to a negative correlation rule, that is, the maximum difference between the street light brightness fluctuation and the baseline value: when the speed reference value is large, such as when the vehicle is traveling at high speed, for example, a reference value of 80 corresponds to a speed of 100km / h, and the brightness fluctuation range is small, such as a fluctuation range of ±5%, to avoid frequent changes in brightness interfering with the driving vision at high speeds; when the speed reference value is small, such as when the vehicle is traveling at low speed, for example, a reference value of 20 corresponds to a speed of 30km / h, and the brightness fluctuation range is large, such as a fluctuation range of ±15%, which can highlight details such as road bumps and small obstacles through subtle changes in brightness, thereby improving safety at low speeds.

[0075] Reference Figure 2 To further improve the scene adaptability of street light adjustment, the specific methods include the following steps:

[0076] After verifying the vehicle's identity and capturing its driving status, key vehicle feature size parameters are calculated based on the extracted first and second vehicle feature images. Combining the pixel proportions and positional relationships of these two types of features in the road image, quantitative indicators reflecting the vehicle's actual size are calculated. For example, the actual width of the vehicle is estimated by the pixel spacing between the left and right headlights in the headlight feature image, combined with the preset focal length and shooting distance parameters of the image sensor. The headlight spacing is usually in a fixed proportion to the vehicle's width; for instance, the headlight spacing of a sedan is approximately 60%-70% of its body width. The pixel height and length of the windshield in the glass feature image are used to help verify the vehicle's height and wheelbase; for example, the windshield height of a truck is usually smaller than that of a sedan, and the windshield length of an SUV is positively correlated with its wheelbase. This combined calculation of feature size parameters effectively avoids errors from single parameters, such as relying solely on headlight spacing, which is easily affected by vehicle modifications, ensuring the accuracy of vehicle size determination.

[0077] The calculated vehicle feature size parameters are matched against a pre-set vehicle size database. This database covers the size classification standards of mainstream vehicle models, such as dividing vehicles into three categories: small cars (length < 4.3 meters, width < 1.8 meters, such as ordinary sedans), mid-size cars (length 4.3-5 meters, width 1.8-1.95 meters, such as SUVs and mid-size sedans), and large cars (length > 5 meters, width > 1.95 meters, such as trucks and large buses), with corresponding size thresholds set for each category. By comparing the vehicle feature size parameters with the database thresholds, the system can quickly determine the size category of the current vehicle. For example, if a vehicle's feature size parameters show a length of 5.2 meters and a width of 2.1 meters, it matches as a large car; if the length is 4.2 meters and the width is 1.75 meters, it matches as a small car.

[0078] After determining the vehicle size category, the system further calculates a size reference value. Using a preset size as a reference, typically the average size of a mid-sized vehicle (e.g., 4.6 meters long and 1.85 meters wide), the actual vehicle size is converted into a quantifiable indicator using the formula: Size Reference Value = (Actual Vehicle Size Parameter / Preset Size Parameter) × 100. For example, compared to the preset length of 4.6 meters, the actual length of a large vehicle (5.2 meters) has a size reference value of approximately 113; the actual length of a small vehicle (4.2 meters) has a size reference value of approximately 91. This reference value directly reflects the difference between the current vehicle size and the benchmark size.

[0079] Following the rule of adjusting the setting distance in a positive correlation with the size reference value, adjust the coverage area of ​​the streetlights in front of the vehicle whose brightness needs to be adjusted: When the size reference value is large, such as 113 for a large vehicle, it indicates that the vehicle is larger and longer, and the driver's blind spot is correspondingly wider. For example, truck drivers have a higher need to observe the road conditions 50 meters ahead of the vehicle. Therefore, the setting distance needs to be extended, such as from the basic 200 meters to 300 meters, to ensure that streetlights within a longer distance in front of the vehicle are within the brightness adjustment range, providing the driver with more observation time and reaction space, and avoiding problems caused by insufficient lighting coverage. Safety hazards include large vehicles failing to detect road obstacles ahead due to short lighting range; when the size reference value is small, such as 91 for small cars, the small vehicle size and small blind spot make the driver's need to observe the road conditions at close range more prominent. In this case, the set distance can be shortened, such as from 200 meters to 150 meters, without the need to adjust the streetlights at excessive distances. This design ensures sufficient lighting along the driving path of small cars, with a range of 150 meters being sufficient to cover their safety observation needs, while avoiding ineffective lighting of streetlights within a range of 200-300 meters, reducing unnecessary power consumption.

[0080] Through this series of steps, the solution achieves precise matching between vehicle size and lighting range, breaking the limitations of the fixed coverage of traditional streetlights: it provides a wider range of safety lighting for large vehicles and optimizes lighting energy consumption for small vehicles, realizing the allocation of lighting resources on demand.

[0081] To ensure that streetlights always match the real-time ambient light requirements, the following steps are included:

[0082] Building upon vehicle recognition and speed determination, the system further analyzes the ambient lighting characteristics of road images. From the pre-processed road images, image segmentation algorithms (such as thresholding and region growing algorithms) accurately extract two types of key image regions: First, shadow images, which are dark areas on the road surface created by vehicles blocking natural light or streetlights, such as long shadows cast by vehicles at midday on a sunny day or localized dark areas created by vehicles blocking streetlights at night. The pixel brightness of these areas is significantly lower than that of unobstructed road surfaces, a major cause of blind spots for drivers. Second, road surface images, which are unobstructed road areas directly illuminated by light, such as the open road surface surrounding vehicle shadows. The pixel brightness of these areas can serve as a benchmark for the current ambient light intensity. Through refined image segmentation, the system effectively eliminates interference factors such as road stains and road marking reflections, ensuring high purity in both the extracted shadow and road surface images, providing reliable data for subsequent pixel value calculations.

[0083] Subsequently, the system enters the pixel value quantization calculation stage: for the extracted shadow image, its shadow pixel value is calculated using a pixel averaging algorithm. Specifically, a noise-free area in the shadow image is selected, such as a 100×100 pixel square area. The brightness values ​​of all pixels within this area are counted and averaged. For grayscale images, the single-channel brightness value is used; for RGB images, the weighted average of the three channels is used. This yields the shadow pixel value, reflecting the overall brightness of the shadow area. The value range is consistent with the image sensor; for example, 0-255 for an 8-bit image. The lower the value, the darker the shadow. Similarly, for the road surface image, an area of ​​the same size is selected, and the road surface pixel value is calculated. This value directly reflects the actual light intensity of the unobstructed road surface in the current environment; the higher the value, the stronger the ambient light. By using averaging calculation instead of single-pixel value, the interference of individual abnormal pixels on the calculation results can be effectively avoided, ensuring the accuracy of shadow pixel values ​​and road surface pixel values. Abnormal pixels include transition pixels at the edge of shadows and reflective pixels from small pebbles on the road surface.

[0084] After acquiring the two types of pixel values, the system calculates the shadow reference value using a preset algorithm. The formula is: Shadow reference value = ((Road surface pixel value - Shadow pixel value) / Pixel value dynamic range) × Reference value scaling factor. The design logic and practical significance of each parameter are as follows:

[0085] The difference between the pixel values ​​of the road surface and the shadow directly reflects the degree of light and dark contrast between the road surface and the shadow area. The larger the difference, the more significant the difference in brightness between the two. For example, on a sunny day at noon, the road surface pixel value is 220 and the shadow pixel value is 80, with a difference of 140, indicating a strong light and dark contrast. The smaller the difference, the closer the brightness of the two. For example, on a cloudy day, the road surface pixel value is 150 and the shadow pixel value is 120, with a difference of 30, indicating a weak light and dark contrast.

[0086] The dynamic range of pixel values ​​is the theoretical maximum and minimum value of the pixel output by the image sensor. For example, for an 8-bit image, it is 255-0=255, and for a 16-bit image, it is 65535-0=65535. Its function is to normalize the difference between brightness and darkness. By dividing by the dynamic range, the difference between brightness and darkness of different hardware devices and different image formats is uniformly mapped to the range of 0-1, eliminating the calculation deviation caused by device differences. For example, the difference between brightness and darkness of 500 in a 10-bit image, after being divided by the dynamic range of 1023, is comparable to the difference between brightness and darkness of 120 in an 8-bit image divided by 255.

[0087] The reference value scaling factor is usually set to 100, which converts the normalized result into an integer range of 0-100. This facilitates system storage and calculation, and also intuitively reflects the size of the shadow reference value. For example, a reference value of 50 represents medium contrast, and a reference value of 80 represents strong contrast, providing a clear quantitative basis for subsequent brightness adjustment.

[0088] For example, in a sunny midday scene, the road surface pixel value is 220, the shadow pixel value is 80, the dynamic range of the pixel value is 255, and the reference value scaling factor is 100. Substituting these values ​​into the formula, we can get the shadow reference value = ((220-80) / 255)×100≈55. In a cloudy scene, the road surface pixel value is 150, the shadow pixel value is 120, and we can calculate the shadow reference value = ((150-120) / 255)×100≈12. The contrast between the light and dark in the two scenes is accurately quantified through the difference in the shadow reference value.

[0089] The initial lighting intensity of streetlights is optimized by adjusting the base value of streetlight brightness according to the positive correlation with the shadow reference value: When the shadow reference value is large, such as 55 on a sunny day, it indicates a strong contrast between light and dark in the environment, and the shadow area is severely underlit, making it easy for drivers to not see road bumps or small obstacles in the shadow. In this case, the base value of brightness needs to be increased, such as from 30% to 60%, to make up for the lighting gap in the shadow area and eliminate blind spots by enhancing the overall light intensity of the streetlight. When the shadow reference value is small, such as 12 on a cloudy day, it indicates that the ambient light is uniform and the difference in brightness between the road surface and the shadow area is small. It is not necessary to have an excessively high base brightness to meet the vision requirements. In this case, the base value of brightness can be reduced, such as from 30% to 20%. This adjustment logic ensures basic lighting for the road surface on cloudy days, with 20% brightness being sufficient for drivers to clearly observe road conditions. It also avoids the energy waste caused by lighting according to the brightness standard for sunny days. At the same time, it can reduce the visual stimulation of strong light on drivers, such as the glare caused by excessive brightness on cloudy days.

[0090] The above solution achieves dynamic adaptation of ambient light and basic brightness values: it not only solves the safety hazard of insufficient lighting in shaded areas on sunny days, but also avoids excessive lighting in low-contrast scenarios such as cloudy days and nights, ensuring that the basic brightness value of streetlights always remains consistent with real-time environmental requirements. This not only improves driving safety and visual comfort, but also further reduces unnecessary energy consumption.

[0091] To address the issues of delayed lighting response when vehicle driving conditions change abruptly, or excessively frequent adjustments when driving at a constant speed, the specific method includes the following steps:

[0092] Based on the calculated speed reference value reflecting the difference between the vehicle's actual speed and the set speed, a fixed set period is established. This period, typically 0.5-1 second, is determined by the image sensor sampling frequency and the vehicle's dynamic characteristics. A period that is too short is susceptible to instantaneous speed fluctuations, while a period that is too long will delay the adjustment response. Within each set period, real-time data of the speed reference value is continuously collected, such as 10 times per second. The rate of change of the speed reference value is then calculated using a difference algorithm. The calculation logic for the rate of change is as follows: using the initial speed reference value within the set period as a baseline, the speed reference value at the end of the period is subtracted from the baseline value, and then divided by the set period duration. This yields the change in the speed reference value per unit time, expressed in reference value units per second. The sign and magnitude of this value directly reflect the trend and severity of the vehicle's speed change. For example, a rate of change of +8 reference value units per second indicates a rapid increase in the speed reference value within one second, corresponding to rapid acceleration; a rate of change of -5 reference value units per second indicates a rapid decrease in the speed reference value, corresponding to rapid deceleration; and a rate of change close to 0 indicates a stable vehicle speed, indicating uniform driving.

[0093] Based on the rate of change of the speed reference value, the minimum adjustment unit for brightness fluctuation is optimized according to the negative correlation adjustment rule. Here, the minimum adjustment unit refers to the minimum brightness increment or decrease for each change during the dynamic adjustment of the streetlight brightness, such as 1% brightness percentage or 5% brightness percentage. The smaller the unit, the higher the adjustment precision and the smoother the brightness change; the larger the unit, the lower the adjustment precision and the more rapid the brightness change. Its adjustment logic is deeply adapted to the vehicle's driving state.

[0094] When the speed reference value changes rapidly, such as during rapid acceleration or deceleration, with the absolute value of the change exceeding 5 reference units per second, the system will increase the minimum adjustment unit for brightness fluctuation, such as from 1% to 5%. In such scenarios, the vehicle's driving state changes rapidly, and the demand for lighting also adjusts quickly accordingly. For example, when a vehicle accelerates rapidly from low speed to high speed, the brightness fluctuation needs to be reduced quickly to ensure stable visibility; when decelerating rapidly from high speed to low speed, the brightness fluctuation needs to be increased quickly to highlight road details. If a small adjustment unit is still used, the brightness adjustment needs to be accumulated multiple times to reach the target value, which can easily lead to the problem of lighting adjustment lagging behind changes in vehicle state. For example, after rapid acceleration, if the brightness fluctuation is not reduced in time, high-frequency changes in brightness and darkness will still exist, interfering with the driver's vision at high speeds. However, by increasing the minimum adjustment unit, the brightness can be quickly adapted to the target state through a single or a few adjustments, ensuring that the lighting is synchronized with the vehicle's driving state and eliminating the potential for lag.

[0095] When the speed reference value changes slowly, such as when the vehicle is traveling at a constant speed and the absolute value of the change is less than 1 reference unit per second, the system will reduce the minimum adjustment unit for brightness fluctuation, such as from 5% to 1%. If a large adjustment unit is used, even if the brightness fluctuation itself is small, a single brightness change will create a noticeable difference in brightness, such as adjusting from 30% brightness to 35% in one go. Frequent adjustments can easily cause driver visual fatigue and even create the illusion of road flickering, affecting driving safety. Small adjustment units can achieve step-by-step fine-tuning of brightness. For example, the brightness change from 30% to 35% can be broken down into five 1% adjustments. Each change is so subtle that it is difficult to detect with the naked eye. This satisfies the needs of slow changes in lighting, such as when the ambient light is slightly reduced during constant speed driving, while maintaining the overall stability of brightness and avoiding interference with the driver's vision.

[0096] In addition, the minimum adjustment unit is smoothly transitioned based on the gradient difference in the rate of change. For example, when the rate of change gradually increases from 1 reference unit / second to 5 reference units / second, the minimum adjustment unit will gradually increase from 1% to 5%, rather than jumping instantaneously. This further avoids brightness jumps caused by abrupt changes in the adjustment unit and ensures that the entire lighting adjustment process is continuous and natural.

[0097] Through this dynamic adjustment mechanism, the solution achieves a precise match between vehicle driving dynamics and lighting adjustment accuracy, ensuring both adjustment response speed when the vehicle's state changes abruptly and visual comfort when the vehicle is driving smoothly.

[0098] To further adapt street lighting to the visual perception needs of drivers under different driving conditions, the specific methods include the following steps:

[0099] Based on the obtained speed comparison values ​​reflecting the difference between the vehicle's actual speed and the set speed, the trend value is calculated. This captures the dynamic characteristics of the speed comparison value within a continuous time interval, typically set to 1-2 seconds, balancing data timeliness and trend stability. Too short an interval is susceptible to instantaneous speed fluctuations, while too long an interval fails to respond promptly to speed changes. The trend value is calculated in two ways, which can be flexibly chosen based on the actual scenario: First, the rate of change, calculated using the formula: Rate of change = (Speed ​​comparison value at the end of the time interval - Speed ​​comparison value at the beginning of the time interval) / Duration of the time interval, yields the magnitude of the speed comparison value change per unit time, intuitively reflecting the speed of change. Second, the amount of change, which is obtained by subtracting the initial value from the final value of the time interval, yielding the absolute change in the speed comparison value within that time period, clearly demonstrating the direction and magnitude of the speed change. Regardless of the method used, the sign and magnitude of the trend value represent vehicle driving status information: if the trend value is positive and large, it means the vehicle speed is rapidly increasing and accelerating; if the trend value is negative and large, it means the vehicle speed is rapidly decreasing and decelerating; if the trend value is close to 0 and within a fixed range, it means the vehicle speed is stable and driving at a constant speed.

[0100] A preset trend value range is established, calibrated through extensive driving vision experiments. This range is typically set to -0.5 to +0.5 reference units per second, with the specific value adjustable based on road speed limits. This range corresponds to the speed variation interval of a vehicle during stable driving. Based on this, the trend values ​​are categorized into three scenarios, each matched with a corresponding color temperature scheme. The color temperature here essentially refers to the color characteristics of the light: low color temperature (2700K-3500K) light is warm yellow, soft, and has strong penetrating power, enhancing the three-dimensionality of objects; medium color temperature (4000K-4500K) light is neutral white, with balanced and natural light, minimizing eye strain; high color temperature (5000K-6500K) light is cool white, bright, and highly clear, highlighting details and contours.

[0101] In terms of specific adjustment logic, the adaptation strategies for the three scenarios are highly compatible with the needs of driving vision:

[0102] When the trend value is within the preset trend value range, such as a trend value of +0.3 reference units / second, and the vehicle is traveling at a constant speed, the system adjusts the street light color temperature to the preset medium color temperature value, typically 4200K. Constant speed driving is the primary driving state for vehicles, requiring drivers to maintain focused vision for extended periods. The medium color temperature neutral white light avoids the visual fatigue that low color temperature warm light may cause after prolonged exposure, and also avoids the visual stimulation caused by high color temperature cold light. The soft and natural light effectively alleviates visual fatigue during long drives, while ensuring clear identification of basic information such as road markings and obstacles, providing a comfortable lighting environment for smooth driving.

[0103] When the trend value exceeds the preset trend value range, such as a trend value of +1.2 reference units / second, and the vehicle accelerates rapidly, the system switches the street light color temperature to a preset low color temperature value, typically 3000K. During acceleration, the relative speed between the vehicle and the road surface increases, significantly enhancing the driver's need to perceive the road's three-dimensional contours. The warm yellow light with a low color temperature has a unique sense of light and shadow, enhancing the contrast between light and shadow on road bumps, potholes, and gravel, making the three-dimensional contours of these potential hazards clearer. For example, under warm light, small bumps on the road surface will form obvious shadows, making it easier for the driver to quickly spot them. Simultaneously, the soft characteristics of warm light can prevent strong glare from interfering with the driver's vision at increased speed, helping the driver maintain a stable judgment of road conditions while driving at high speeds.

[0104] When the trend value is less than the preset trend value range, such as a trend value of -1.0 reference unit / second, the vehicle decelerates rapidly, and the system adjusts the street light color temperature to the preset high color temperature value, usually 5500K. Deceleration often occurs in scenarios requiring careful observation, such as approaching intersections, avoiding pedestrians, or encountering obstacles. Drivers at these times need to focus on identifying road surface details and the surrounding environment. High color temperature cool white light offers high brightness and clarity, significantly improving the visual visibility of the road surface, clearly revealing small cracks and scattered debris, while also making the outlines of pedestrians and non-motorized vehicles more distinct. Furthermore, the strong penetrating power of cool light can expand the driver's effective field of vision at night or in low-visibility environments, allowing more reaction time for judgment and operation during deceleration, thus reducing safety risks.

[0105] The above solution achieves deep adaptation between vehicle driving status and light color temperature, making streetlights no longer a single fixed light source, but an intelligent visual aid that can dynamically adjust according to the driver's real-time needs; it ensures comfort when driving smoothly and enhances safety when speed changes, further improving the humanization and practicality of smart city street lighting, and providing strong support for driving safety at night and in complex road conditions.

[0106] To ensure that the street light color temperature adjustment more accurately matches the subtle differences in vehicle speed, the following steps are included:

[0107] After obtaining the speed reference value change trend, the system first introduces a preset reference trend value. This value is set based on common driving conditions on urban roads, usually corresponding to the ideal trend value when the vehicle is driving smoothly. For example, a reference value of 0.2 units / second represents slow speed change and stable driving conditions, serving as a reference for judging the degree to which speed changes deviate from the stable benchmark. The change trend calculation value is calculated through a preset algorithm, with the formula: Change trend calculation value = |Change trend value - Reference trend value|. That is, through absolute value calculation, the deviation of the speed trend from the benchmark is converted into a non-negative quantitative index: If the change trend value is close to the reference trend value, such as a change trend value of 0.3 and a reference trend value of 0.2, the calculated value is 0.1 smaller, representing that the speed change is close to a stable state; if the change trend value deviates greatly from the reference trend value, such as a change trend value of 1.5 and a reference trend value of 0.2, the calculated value is 1.3 larger, representing drastic speed changes, rapid acceleration or deceleration.

[0108] Based on this, the system adapts the street light color temperature according to the inverse correlation adjustment rules; here, the brightness of the color temperature corresponds to the color temperature value: high color temperature (5000K-6500K) light is cool white, visually brighter, and has high clarity; low color temperature (2700K-3500K) light is warm yellow, visually darker, and has high softness. The specific adjustment logic is closely aligned with driving needs: When the calculated value of the change trend is small, the speed change is gradual, such as driving at a constant speed, indicating that the driver does not need to frequently adjust their attention. At this time, the color temperature is increased (e.g., 6000K), and the high clarity of cool white light makes details such as road markings and small obstacles easier to identify, while avoiding the visual drowsiness that warm light may cause. When the calculated value of the change trend is large, the speed change is drastic, such as rapid acceleration and deceleration, and the driver needs to concentrate highly to deal with changes in road conditions. At this time, the color temperature is decreased (e.g., 3000K), and the soft characteristics of warm yellow light reduce the stimulation of strong light on the eyes and alleviate the visual pressure of rapid changes. For example, during rapid deceleration, warm light can avoid the sudden pupil constriction caused by direct exposure to cool light, thus improving visual comfort.

[0109] Furthermore, this adjustment logic features gradient adaptation: as the calculated trend value gradually increases from 0.1 to 1.5, the color temperature smoothly decreases from 6000K to 3000K, rather than changing abruptly, ensuring a natural transition in light color and preventing sudden color temperature changes from interfering with the driver's vision. For example, as the vehicle accelerates slowly from a constant speed (calculated value 0.1, color temperature 6000K) to a slow acceleration (calculated value 0.5, color temperature 5000K) and then to rapid acceleration (calculated value 1.5, color temperature 3000K), the color temperature gradually adjusts with the speed change trend, always remaining synchronized with the driver's visual needs.

[0110] Through this series of designs, the solution achieves precise control over the range of speed changes and color temperature adaptation. This avoids the limitations of fixed color temperature in dealing with diverse driving conditions, and improves the continuity and comfort of driving vision through gradient adjustment, further enhancing the intelligence and humanization of smart city street lighting.

[0111] To address the issue of the fixed-interval street light quantity being insufficient to accommodate varying vehicle types and road conditions, the specific method includes the following steps:

[0112] Building upon vehicle recognition and speed determination, the system further integrates LiDAR sensing data. Deployed on streetlights or roadsides, LiDAR devices offer high ranging accuracy and a wide scanning angle, typically covering a range of 100-200 meters. They emit continuous scanning light rays around vehicles, which reflect off the vehicle body, road surface, and surrounding environment, forming a laser-sensed image containing distance and contour information. Compared to traditional image sensors, laser-sensed images are unaffected by lighting conditions, providing accurate imaging even at night or in rainy weather, and can directly output coverage area data in three-dimensional space.

[0113] Regional analysis is performed on the laser-sensing images to calculate the coverage area. This coverage area specifically refers to the spatial range within which the LiDAR scanning light can effectively capture details of the vehicle and its surrounding environment, typically measured in square meters. The calculation dimensions include the lateral coverage width (e.g., the sum of the distances from both sides of the vehicle to the scanning boundary) and the longitudinal coverage length (e.g., the length from the front of the vehicle to the furthest point of the scan). For example, when a large truck travels on an open main road, the LiDAR scan is unobstructed, covering the road surface and 2-3 meters on both sides laterally, and up to 150 meters in front longitudinally, resulting in a coverage area of ​​over 300 square meters. However, when a small car travels on a narrow side road, with guardrails or roadside trees obstructing the view, the lateral coverage is only 1-1.5 meters, and the longitudinal coverage is limited to 80 meters in front due to building obstructions, potentially resulting in a coverage area of ​​less than 100 square meters. Using the point cloud data from the laser-sensing images, each reflection point represents a spatial coordinate, automatically calculating the actual area of ​​the coverage region and accurately quantifying the effective perception range in the current scene.

[0114] After obtaining the coverage area size, the system optimizes the number of streetlights in each interval of brightness interval changes according to a positive correlation adjustment rule. Here, brightness interval refers to dividing the streetlights within a set distance in front of the vehicle into multiple continuous intervals, such as every 50 meters. Streetlights within each interval simultaneously execute the same brightness adjustment command, such as simultaneously increasing to 80% brightness. The number of streetlights in each interval determines the coverage density of a single lighting zone; the more streetlights, the higher the lighting density and the more continuous the coverage. Its adjustment logic closely matches the scenario requirements.

[0115] When the coverage area is large, such as for large vehicles or open roads, it means that the area around the vehicle needs to be illuminated more widely. Large vehicles have long bodies and large blind spots, requiring a longer and wider illumination range ahead to identify road conditions in advance. Open roads have no obstructions, and light travels a long distance, covering a larger area. In this case, the system will increase the number of streetlights in each interval, such as increasing from 2 streetlights per interval to 4 streetlights per interval. This allows a single lighting interval to cover a wider road surface and a longer distance, avoiding gaps in illumination due to insufficient streetlights. If there are weak light areas between two intervals, it ensures continuous and sufficient lighting along the vehicle's path, eliminating blind spots caused by incomplete lighting coverage.

[0116] When the coverage area is small, such as in areas with small vehicles or narrow side roads, the vehicles themselves are small in size and have small blind spots, resulting in low demand for large-area lighting. Narrow side roads have limited width, and too many streetlights operating simultaneously can cause light overlap. In such cases, the system reduces the number of streetlights per interval, for example, from four to two per interval, retaining only the number needed for basic lighting. This avoids over-illumination of small vehicles by using the same number of streetlights to cover large vehicles, such as four streetlights simultaneously shining brightly on a narrow road, which would create light redundancy. It also reduces unnecessary energy consumption and avoids glare caused by overlapping strong light, such as multiple streetlights shining directly into the driver's line of sight, thus improving driving visual comfort.

[0117] In addition, when a vehicle moves from an open road into a narrow side road, the coverage area decreases, and the laser sensing image captures the change in coverage area in real time, reducing the number of streetlights in each interval accordingly. When the vehicle changes from a small car to a large truck, the LiDAR identifies the vehicle type to assist in verifying the coverage area. As the coverage area increases, the number of streetlights will also increase in real time to ensure that the lighting adjustment is always synchronized with the current scene.

[0118] The above solution allows lighting resources to be flexibly allocated according to differences in vehicle type and road environment, ensuring lighting continuity in large-scale scenarios while avoiding energy waste in small-scale scenarios.

[0119] To more accurately adapt street light brightness to the needs of environmental detail perception, the specific methods include the following steps:

[0120] Based on the acquired laser-sensing image, the scanning resolution of a LiDAR is essentially the density of laser scanning points per unit space, such as 50 scanning points per square meter or 10 scanning points per square meter. Higher resolution means a stronger ability of the laser light to capture environmental details, such as clearly identifying small cracks in the road surface and low obstacles along the roadside. Lower resolution means weaker detail capture, only able to identify macroscopic targets such as vehicles and large obstacles. The resolution is usually related to environmental complexity and the LiDAR's operating status: for example, when driving in complex road conditions such as around commercial areas with many pedestrians, non-motorized vehicles, and abundant road facilities, the LiDAR will automatically increase its scanning resolution to capture more details; while driving on open suburban roads with no complex interference and a smooth road surface, the resolution will be appropriately reduced to decrease data processing load. The system can directly read the real-time resolution value of the current scanning light through the hardware parameter feedback of the LiDAR, providing a quantitative basis for subsequent adjustments.

[0121] The system dynamically optimizes the brightness increment as the streetlight increases brightness from its base brightness to the target brightness, following a positive correlation adjustment rule. The brightness increment refers to the magnitude of the increase in brightness with each adjustment as the streetlight progresses from its base brightness to the target brightness. For example, a 5% increase or a 2% increase results in a faster brightness increase and a more significant final high-brightness state; a smaller increment results in a smoother brightness increase and a more gentle final light intensity. Its adjustment logic closely matches the environmental detail requirements reflected by the resolution.

[0122] When the scanning light resolution is high, such as 50 scanning points per square meter, corresponding to complex road conditions, it indicates that there are many detailed targets in the current environment that require the driver's attention, such as temporarily parked shared bicycles at intersections or bumps left by road repairs. In this case, it is necessary to increase the brightness increment, such as from the basic increment of 2% to 5%. On the one hand, a larger brightness increment allows the streetlights to quickly reach a higher illumination intensity, enhancing the illumination coverage of key detail areas. For example, high brightness can make road cracks and surrounding road surfaces more obvious, making it easier for drivers to spot them in time. On the other hand, at high resolution, drivers have a higher demand for detail perception. Sufficient brightness increment can prevent details from being hidden due to insufficient light. If the brightness increase is slow with low increment, it is easy to miss risk points in complex road conditions due to insufficient light, thus providing drivers with a clear field of vision for observing details.

[0123] When the scanning light resolution is low, such as 10 scanning points per square meter, corresponding to simple road conditions, it indicates that the current environment does not have many complex details. Drivers only need to focus on macroscopic road conditions, such as vehicles ahead and road direction. In this case, the brightness increment should be reduced, such as from 5% to 2%. This ensures basic lighting needs are met, and the gradual accumulation of 2% increments is sufficient to keep the road surface clearly visible, while avoiding excessive lighting caused by high increments. For example, if a large increment of 5% is still used in an open road section, the streetlights will quickly reach full brightness, resulting in light redundancy. If there are no details on the road surface to see, the strong light will cause visual fatigue for drivers and also cause unnecessary energy consumption. The gradual brightness increase with small increments can minimize energy waste while meeting the visibility needs, which is in line with the energy-saving concept of smart cities.

[0124] Furthermore, as the vehicle transitions from complex to simple road conditions, the scanning resolution gradually decreases, and the brightness increment is smoothly adjusted from large to small, rather than abruptly changing. For example, when the resolution decreases from 50 pixels per square meter to 30 pixels per square meter, the increment decreases from 5% to 3%; when the resolution further decreases to 10 pixels per square meter, the increment decreases to 2%. This transition avoids sudden changes in lighting caused by abrupt increments, ensuring the continuity and comfort of the driver's vision.

[0125] The above solution achieves a precise match between the accuracy of environmental detail perception and the degree of brightness improvement: it solves the problems of insufficient brightness and difficulty in distinguishing details in high-resolution complex scenes, while avoiding the pain points of excessive lighting and energy waste in low-resolution simple scenes.

[0126] To more accurately adapt the reduction of street light brightness to the dynamic changes in vehicle movement, the specific methods include the following steps:

[0127] Based on the acquired laser-sensing image, the system focuses on extracting the scanning light frequency. The scanning light frequency of the LiDAR refers to the number of times the laser beam scans the same area per unit time, such as 30 times per second or 10 times per second. A higher frequency means the laser captures dynamic changes in the vehicle and its surrounding environment more promptly, allowing for real-time tracking of rapid vehicle movement and sudden changes in direction. A lower frequency results in a smoother dynamic capture rhythm, suitable for scenarios where the vehicle is moving at a constant speed and the environment remains relatively stable. The frequency is typically related to the vehicle's motion state: when a vehicle quickly moves away from the streetlight's monitoring range, such as accelerating away after overtaking on a highway, the LiDAR automatically increases the scanning frequency to accurately track its position change; when the vehicle is moving slowly and steadily on a residential road, the frequency remains lower to reduce redundant calculations. The system can directly obtain the quantized value of the current scanning light frequency through the LiDAR's real-time operating parameters, providing a dynamic basis for brightness adjustment.

[0128] Following a positive correlation adjustment rule, the brightness reduction during dimming is dynamically optimized. This brightness reduction refers to the magnitude of the brightness decrease during each adjustment as the streetlight transitions from high brightness to baseline brightness. For example, a reduction of 8% or 3% brightness per adjustment results in a faster brightness decay and a quicker return to a low-energy state; a smaller reduction results in a smoother brightness decay and a more natural light transition. This adjustment logic closely aligns with the vehicle's dynamic characteristics reflected by the frequency.

[0129] When the scanning frequency is high, such as 30 times per second, it corresponds to vehicles moving away rapidly or undergoing drastic dynamic changes. This indicates that the relative position of the vehicle and the streetlight is changing rapidly, and the original illuminated area no longer needs to maintain high brightness. The original illuminated area could be a section of road that a vehicle has just passed through. In this case, the system will increase the brightness reduction, for example, from the base reduction of 3% to 8%. Through a significant and rapid brightness reduction, the streetlight, which has moved away from the vehicle's needs, quickly returns to a low-brightness state. For example, after a vehicle drives away at 80 km / h, a high-frequency scan can confirm in real time that it has exceeded the lighting requirements. The large reduction of 8% allows the streetlight to drop from 80% brightness to 30% of the base value within 2-3 adjustments, avoiding energy waste caused by the streetlight remaining bright for a long time after the vehicle has moved away, while also reducing light pollution to surrounding residents.

[0130] When the scanning frequency is low, such as 10 times per second, it corresponds to a vehicle moving at a constant speed or stationary, indicating that the vehicle is dynamically stable and the rhythm of changes in the illuminated area is gradual. At this time, the system will reduce the brightness reduction when the brightness dims, such as from 8% to 3%, to achieve a smooth transition through small, gradual brightness reductions. For example, when a vehicle passes through a road segment at a constant speed of 20 km / h, the low-frequency scanning reflects its slow positional change. A small reduction of 3% can gradually reduce the street light brightness from 80% to 30%, requiring 17 adjustments to avoid a sudden drop in brightness due to a large reduction, such as a sudden drop from 80% to 50%. This smooth transition can reduce the interference of sudden changes in light brightness on the driver's vision, especially at night, when a sudden dimming of strong light can easily cause pupil constriction, resulting in temporary visual blurring, thus ensuring visual comfort and continuity during driving.

[0131] In addition, when the vehicle changes from high speed to constant speed, the scanning frequency drops from 30 times / second to 10 times / second, and the brightness reduction will smoothly transition from 8% to 3% in sync, rather than switching instantaneously. For example, when the frequency drops to 20 times / second, the reduction is first adjusted to 5%; when the frequency drops further to 15 times / second, the reduction is adjusted to 4%, ensuring that the brightness decay rhythm is always synchronized with the dynamic changes of the vehicle and avoiding the adjustment action being out of sync with actual needs.

[0132] The above solution achieves a precise match between vehicle dynamic capture efficiency and brightness attenuation: it solves the energy waste caused by lighting lag in high-frequency scenarios and avoids visual discomfort caused by sudden changes in brightness in low-frequency scenarios, allowing the streetlight brightness reduction process to respond quickly to dynamic changes and smoothly adapt to stable conditions.

[0133] This application also discloses a smart city street lighting adjustment system, including a processor, which executes the steps of the smart city street lighting adjustment method as described in any of the above embodiments.

[0134] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for adjusting street lighting in a smart city, characterized in that, Includes the following steps: The road image is acquired using an image sensor installed on the street lamp, and a first vehicle feature image and a second vehicle feature image are extracted from the road image; the first vehicle feature image is a headlight image, and the second vehicle feature image is a glass image; A first image matching value is calculated based on the first vehicle feature image and a preset first feature template; a second image matching value is calculated based on the second vehicle feature image and a preset second feature template; a feature position matching value is calculated based on the first vehicle feature image and the second vehicle feature image; wherein, the pixel spacing and angle deviation of the first vehicle feature image and the second vehicle feature image in the road image are calculated, and the feature position matching value is output. The vehicle matching value is calculated based on the first image matching value, the second image matching value, and the feature position matching value. If the vehicle matching value is greater than the preset matching value, the vehicle direction is calculated based on the first vehicle feature image and the second vehicle image, and the first continuous timestamp of the first vehicle feature image and the second continuous timestamp of the second vehicle feature image are obtained. A first speed is calculated based on the displacement of the first vehicle feature image on the road image and a first consecutive timestamp; a second speed is calculated based on the displacement of the second vehicle feature image on the road image and a second consecutive timestamp; a characteristic speed is calculated based on the first and second speeds; if the characteristic speed is greater than a preset speed, a speed comparison value is calculated based on the characteristic speed and the preset speed, and the brightness interval of the streetlights within a preset distance in front of the current vehicle is controlled. The brightness of the streetlights is adjusted inversely based on their distance from the vehicle; the farther the streetlight is, the lower its brightness; the closer the streetlight is, the brighter its brightness. The brightness fluctuation amplitude is adjusted based on the negative correlation between the speed reference value and the brightness interval. The larger the speed reference value, the smaller the brightness fluctuation amplitude, and vice versa.

2. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Calculate the vehicle feature size parameters based on the first vehicle feature image and the second vehicle feature image; The vehicle dimensions are matched based on the vehicle characteristic dimension parameters; The size reference value is calculated based on the vehicle size and the preset size. The setting distance is adjusted according to the size reference value. The larger the size reference value, the longer the setting distance, and the smaller the size reference value, the shorter the setting distance.

3. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Extract shadow and road surface images from road images; Calculate the shadow pixel values ​​of the shadow image and the road surface pixel values ​​of the road surface image; The shadow reference value is calculated based on the shadow pixel value and the road surface pixel value according to a preset algorithm. The shadow reference value = ((road surface pixel value - shadow pixel value) / pixel value dynamic range) × reference value scaling factor; the pixel value dynamic range is the difference between the theoretical maximum and minimum values ​​of the pixel values ​​output by the image sensor. The base brightness value of the street light is adjusted according to the positive correlation with the shadow reference value. The larger the shadow reference value, the higher the base brightness value of the street light; the smaller the shadow reference value, the lower the base brightness value of the street light.

4. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Calculate the rate of change of the speed reference value within a set period; The minimum adjustment unit for brightness fluctuation amplitude is adjusted based on the negative correlation between the rate of change and the rate of change. The faster the rate of change, the larger the minimum unit, and the slower the rate of change, the smaller the minimum unit.

5. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Calculate the trend value of the speed reference value, where the trend value is the rate or amount of change of the speed reference value over a continuous time interval, and adjust the color temperature of the street light according to the trend value. If the trend value is greater than the value within the preset trend value range, the color temperature of the street light will be the preset low color temperature value. If the trend value is within the preset trend value range, the color temperature of the street light is the preset medium color temperature value. If the trend value is less than the value within the preset trend value range, then the color temperature of the street light is the preset high color temperature value.

6. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Calculate the trend value of the speed reference value, where the trend value is the rate or amount of change of the speed reference value over a continuous time interval; The calculated value of the changing trend is determined based on the changing trend value and the preset reference trend value. The color temperature of the streetlights is adjusted based on the inverse correlation of the calculated value of the trend of change. The smaller the calculated value of the trend of change, the brighter the color temperature of the streetlights. The larger the calculated value of the trend change, the darker the color temperature of the street light.

7. The smart city street light lighting adjustment method according to claim 1, characterized in that, The method also includes the following steps: Acquire laser-sensing images of the light-shaped beams scanned by the vehicle's LiDAR; The size of the coverage area is calculated based on the laser-sensed image; The number of streetlights in each interval is adjusted according to the positive correlation between the coverage area and the brightness interval. The larger the coverage area, the more streetlights are in each interval; the smaller the coverage area, the fewer streetlights are in each interval.

8. The smart city street light lighting adjustment method according to claim 7, characterized in that, The method also includes the following steps: Obtain the scanning light resolution of the laser-sensing image; The brightness increment is adjusted according to the positive correlation between the scanning light resolution and the brightness increase; the higher the scanning light resolution, the greater the brightness increment. The lower the resolution of the scanning light, the smaller the brightness increment when the light becomes brighter.

9. The smart city street light lighting adjustment method according to claim 7, characterized in that, The method also includes the following steps: Acquire the scanning light frequency of the laser-sensing image; The brightness reduction during dimming is adjusted according to the positive correlation between the scanning light frequency and the brightness frequency; the higher the scanning light frequency, the greater the brightness reduction during dimming. The lower the scanning light frequency, the smaller the brightness reduction when the brightness dims.

10. A smart city street light adjustment system, characterized in that, The system includes a processor that performs the steps of the smart city street lighting adjustment method as described in any one of claims 1-9.

Images