A tunnel lighting control method and system based on video monitoring

Abstract

The present application belongs to the field of lighting control technology and discloses a tunnel lighting control method and system based on video surveillance. The method comprises: identifying lamps, lanes, and vehicles based on video images, and then determining the lane type; dividing the tunnel into multiple detection areas and calculating a conversion ratio factor between pixel distance and actual distance; determining the vehicle speed and the arrival time of the vehicle at the next lamp using a target tracking algorithm based on the vehicle recognition result and the conversion ratio factor; calculating the road space occupancy rate of each detection area based on the vehicle recognition result, the conversion ratio factor, and the lane type, to determine the target color temperature and corresponding target power of the lamps in each detection area, and controlling the operation of multiple lamps in front of the vehicle accordingly. The method can realize real-time and refined lighting control based on actual traffic conditions, effectively save energy, and improve driving safety and energy utilization efficiency.

Description

Technical Field

[0001] The present application relates to the field of lighting control technology, and in particular to a tunnel lighting control method and system based on video surveillance. Background Art

[0002] Tunnels are critical transportation facilities, and their internal lighting environment plays a vital role in driving safety. Traditional tunnel lighting systems have limitations. They often use a unified on / off control method, switching all lamps in the tunnel on and off at a preset on / off time. This makes it difficult to make real-time adjustments based on actual traffic conditions and environmental changes. For example, all lamps are often left illuminated when no vehicles are passing, resulting in a waste of electricity. Furthermore, existing traditional tunnel lighting systems generally only control the on / off status and brightness of lamps and do not adjust color temperature based on actual traffic conditions. In reality, different color temperatures have different visual effects on the driver, and different color temperatures require different power under the same luminous flux conditions. Using a unified color temperature for lighting is not conducive to improving driving safety or energy efficiency.

[0003] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0004] The purpose of this application is to provide a tunnel lighting control method and system based on video surveillance, which can realize real-time and refined lighting control based on actual traffic conditions, effectively save energy, and improve driving safety and energy utilization efficiency.

[0005] In a first aspect, the present application provides a tunnel lighting control method based on video surveillance, comprising the steps of:

[0006] A1. Capture video images inside the tunnel to identify lamps, lanes, and vehicles.

[0007] A2. Determine the lane type based on the lane recognition results; the lane type includes straight or curved;

[0008] A3. Divide the tunnel into multiple detection areas based on the preset lamp spacing and calculate the conversion factor between the pixel distance and the actual distance based on the lamp recognition results, lane type, and the lamp spacing;

[0009] A4. Based on the vehicle identification results and the conversion scale factor, a target tracking algorithm is used to determine the vehicle's position in each detection area, calculate the vehicle's speed, and predict the vehicle's arrival time at the next light fixture.

[0010] A5. Calculate the road space occupancy rate for each detection area based on the vehicle identification results, the conversion scaling factor, and the lane type to determine the target color temperature and corresponding target power of the lamps in each detection area;

[0011] A6. Control the on / off, operating color temperature, and operating power of the plurality of lamps in front of the vehicle based on the speed, the lamp spacing, the arrival time, the target color temperature, and the target power.

[0012] Preferably, step A1 includes:

[0013] A101 obtains the original video image in the tunnel, performs preprocessing, obtains the preprocessed video image; the preprocessing includes grayscale conversion processing for converting the color image into a grayscale image, Gaussian filtering processing and histogram equalization processing;

[0014] A102. Use an object detection algorithm to identify lamps and vehicles in preprocessed video images.

[0015] A103. Use the Canny algorithm to identify lanes in the preprocessed video image.

[0016] Preferably, step A2 includes:

[0017] A201. Detect the identified lane using a Hough transform to determine whether the lane is a straight line.

[0018] A202. If no straight line is detected, determine that the driving lane is a curve, and use a curve fitting method to fit the curve equation of the driving lane.

[0019] Preferably, step A3 includes:

[0020] A301. Divide the tunnel lengthwise into multiple detection zones based on preset lamp spacing.

[0021] A302. Extract the pixel coordinates of the lamp based on the lamp identification result;

[0022] A303. If the lane is a straight line, then the average straight line pixel distance between adjacent lamps is calculated based on the pixel coordinates of the lamps as the lamp pixel spacing;

[0023] A304. If the lane is a curve, then based on the pixel coordinates of the lamps and the curve equation of the lane, calculate the average pixel distance between adjacent lamps along the curve equation as the lamp pixel spacing;

[0024] A305. Calculate a conversion ratio factor between the pixel distance and the actual distance based on the lamp pixel pitch and the lamp pitch.

[0025] Preferably, step A4 includes:

[0026] A401. Based on the vehicle identification results and the conversion scale factor, the target tracking algorithm is used to track the position of the vehicle in each detection area to obtain the vehicle's movement trajectory;

[0027] A402. Calculate the vehicle speed based on the moving trajectory;

[0028] A403. Calculate the arrival time of the vehicle at the next light fixture based on the vehicle's current position and the position of the next light fixture, combined with the vehicle's speed.

[0029] Preferably, step A5 includes:

[0030] A501. Calculate the road area of ​​each detection area based on lane type and preset road width;

[0031] A502. Calculate the actual area of ​​the vehicle's target detection frame based on the pixel size of the vehicle's target detection frame and the conversion scale factor. Using the actual area of ​​the vehicle's target detection frame as the vehicle's occupied area, calculate the sum of the occupied areas of all vehicles in each detection area.

[0032] A503. Divide the total occupied area by the corresponding road area to obtain the road space occupancy rate of each detection area;

[0033] A504. Compare the road space occupancy rate with the preset occupancy rate threshold to determine the target color temperature and corresponding target power of the lamps in each detection area.

[0034] Preferably, step A504 includes:

[0035] If the road space occupancy rate is greater than a first occupancy rate threshold, determining the target color temperature as a first preset color temperature;

[0036] If the road space occupancy rate is less than the second occupancy rate threshold, the target color temperature is determined to be the second preset color temperature; if the first occupancy rate threshold is greater than the second occupancy rate threshold, the first preset color temperature is greater than the second preset color temperature;

[0037] If the road space occupancy rate is between the second occupancy rate threshold and the first occupancy rate threshold, determining the target color temperature based on a linear interpolation operation, or determining the target color temperature as a third preset color temperature; the third preset color temperature is greater than the second preset color temperature and less than the first preset color temperature;

[0038] Obtain the light efficiency parameters corresponding to the determined target color temperature and the luminous flux target value required by the tunnel lighting design;

[0039] The target power corresponding to the target color temperature is obtained by dividing the target luminous flux value by the light efficiency parameter.

[0040] Preferably, step A6 includes:

[0041] A601. Based on the speed and the lamp spacing, determine several lamps in front of the vehicle as the target lamp group;

[0042] A602. Determine the on and off time of the target lighting group based on the arrival time;

[0043] A603. Turn on the target lamp group at the start time, and set the target lamp group's operating color temperature to the corresponding target color temperature, and its operating power to the corresponding target power;

[0044] A604. If a vehicle is detected arriving at the target lamp group before the closing time, a delay time is calculated based on the vehicle's speed to delay closing the target lamp group; if the vehicle does not arrive at the target lamp group within the closing time, the target lamp group is closed.

[0045] Preferably, after step A603 and before step A604, the method further includes:

[0046] A605. Use an image detection method to detect whether the lamps in the target lamp group are successfully turned on, and issue an abnormal warning if they are not successfully turned on.

[0047] In a second aspect, the present application provides a tunnel lighting control system based on video surveillance, comprising a plurality of lamps arranged at intervals along the tunnel, a number of cameras, and a control terminal;

[0048] The camera is used to collect video images in the tunnel and send them to the control terminal;

[0049] The control terminal is used to execute:

[0050] Identify lamps, lanes, and vehicles based on the video image;

[0051] Determine the lane type based on the lane recognition result; the lane type includes straight line or curve;

[0052] Dividing the tunnel into multiple detection areas according to a preset lamp spacing, and calculating a conversion factor between pixel distance and actual distance based on the lamp recognition results, lane type, and the lamp spacing;

[0053] Based on the vehicle identification result and the conversion scale factor, a target tracking algorithm is used to determine the position of the vehicle in each detection area, thereby calculating the vehicle's speed and predicting the vehicle's arrival time at the next light fixture;

[0054] Calculating the road space occupancy rate of each detection area based on the vehicle recognition result, the conversion scale factor, and the lane type to determine the target color temperature and corresponding target power of the lamps in each detection area;

[0055] According to the speed, the lamp spacing, the arrival time, the target color temperature, and the target power, the switching, operating color temperature, and operating power of the plurality of lamps in front of the vehicle are controlled.

[0056] Beneficial effects: The tunnel lighting control method and system based on video surveillance provided in this application obtain traffic information through video surveillance, and dynamically adjust the switch, color temperature and power of lamps based on this information, thereby realizing on-demand lighting. It can realize real-time and refined lighting control based on actual traffic conditions, effectively save energy, and improve driving safety and energy utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flowchart of a tunnel lighting control method based on video surveillance provided in an embodiment of the present application.

[0058] Figure 2 A schematic diagram of the structure of a tunnel lighting control system based on video surveillance provided in an embodiment of the present application.

[0059] Explanation of the numbers: 1. Lighting fixture; 2. Camera; 3. Control terminal. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0061] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0062] Please refer to Figure 1In some embodiments of the present application, a tunnel lighting control method based on video surveillance includes the following steps:

[0063] A1. Capture video images inside the tunnel to identify lamps, lanes, and vehicles.

[0064] A2. Determine the lane type based on the lane recognition results; the lane type includes straight or curved;

[0065] A3. Divide the tunnel into multiple detection zones based on preset lamp spacing. Calculate the conversion factor between pixel distance and actual distance based on lamp recognition results, lane type, and lamp spacing.

[0066] A4. Based on the vehicle identification results and the conversion scale factor, the target tracking algorithm is used to determine the vehicle's position in each detection area. This is used to calculate the vehicle's speed and predict the vehicle's arrival time at the next light.

[0067] A5. Calculate the road space occupancy rate for each detection area based on the vehicle identification results, conversion scale factor, and lane type to determine the target color temperature and corresponding target power of the lamps in each detection area.

[0068] A6. Controls the on / off, operating color temperature, and operating power of several lamps in front of the vehicle based on speed, lamp spacing, arrival time, target color temperature, and target power.

[0069] The core of this application is to use video surveillance to obtain real-time traffic information, and based on this information, to achieve on-demand, local, multi-parameter (on / off, color temperature, power) control of tunnel lamps, thereby achieving energy conservation and providing a suitable lighting environment.

[0070] The working process and principle of this application is to obtain real-time video images in the tunnel and dynamically and regionally control the tunnel lighting based on the image analysis results.

[0071] Specifically, step A1 acquires video images from within the tunnel, which serve as the basis for subsequent processing. These images are analyzed to identify lamps, lanes, and vehicles. Lamp identification provides location information for lighting equipment, lane identification provides information on the road's geometry, and vehicle identification provides dynamic information about traffic participants.

[0072] Based on the lane recognition results, step A2 determines the lane type. The lane type is classified as straight or curved, which provides a geometric basis for subsequent accurate distance calculation and area division.

[0073] Step A3 divides the tunnel length into multiple detection zones based on the spacing between lamps (which can be determined based on tunnel lighting design requirements or pre-measured data). Each zone corresponds to one or a group of lamps, achieving regionalized lighting control. Based on the lamp identification results, lane type, and lamp spacing, a conversion factor between pixel distance and actual distance is calculated. The lamp identification results determine the pixel positions of the lamps. Based on the lane type, the pixel distance between the lamps is calculated using a calculation method. The ratio of the lamp spacing to the pixel distance is then calculated to obtain the conversion factor between pixel distance and actual distance, providing support for the subsequent accurate calculation of speed and road space occupancy.

[0074] Based on the vehicle identification results, step A4 uses a target tracking algorithm to determine the vehicle's position within each detection area. By continuously tracking the vehicle's position changes, the vehicle's real-time speed is calculated. Furthermore, based on the vehicle's current position and the location of the next lamp, combined with the vehicle's speed, the vehicle's arrival time at the next lamp is predicted. This predicted arrival time enables the system to proactively adjust lighting conditions.

[0075] Based on the vehicle identification results, the conversion scaling factor, and the lane type, step A5 calculates the road space occupancy rate. Using the identified vehicle information, the proportion of space occupied by the vehicle on the road is estimated. Combined with the lane type (which influences road area calculation), the road space occupancy rate is calculated to reflect the level of traffic congestion. Based on the road space occupancy rate, an appropriate target color temperature and corresponding target power are determined. By linking lighting parameters to traffic density, the lighting can adapt to varying traffic conditions. For example, when the road is congested, the color temperature is increased to enhance the driver's attention and help them better cope with the congested conditions. When the road is clear, the color temperature is lowered to provide a more comfortable visual experience for the driver and reduce visual fatigue during long drives.

[0076] Finally, based on the predicted arrival time of the vehicle at the next lamp, the determined target color temperature, and the corresponding target power, step A6 identifies several lamps in front of the vehicle as control targets and controls their on / off states, operating color temperature, and operating power. This is the final execution step of the method, converting the information obtained from the previous analysis and calculations into specific control instructions, achieving on-demand lighting and intelligent dimming and color adjustment.

[0077] Through the above solution, this application dynamically adjusts lighting based on real-time tunnel traffic conditions, avoiding unnecessary lighting during periods of low traffic or no traffic, thus reducing energy consumption. Furthermore, by adjusting the color temperature and power of the lighting based on traffic density, it provides lighting effects that match the actual traffic environment, improving driving safety and driver visual comfort. This solves the problem of traditional tunnel lighting systems being unable to adjust lighting in real time according to actual traffic conditions, resulting in energy waste and poor lighting effects.

[0078] In some embodiments, step A1 comprises:

[0079] A101. Obtaining the original video image in the tunnel, performing preprocessing to obtain a preprocessed video image; the preprocessing includes grayscale conversion processing, Gaussian filtering, and histogram equalization processing for converting the color image into a grayscale image;

[0080] A102. Use an object detection algorithm to identify lamps and vehicles in preprocessed video images.

[0081] A103. Use the Canny algorithm to identify lanes in the preprocessed video image.

[0082] Among them, in step A101, grayscale conversion converts the color image into a grayscale image, which simplifies the image data, reduces computational complexity, and eliminates possible interference caused by color information, so that subsequent feature extraction and recognition are more focused on the structure and brightness information of the image. Gaussian filtering is used to smooth the image, remove noise in the image, reduce the impact of noise on the recognition results, and improve the stability of recognition. Histogram equalization is used to enhance the contrast of the image, especially in tunnel environments with uneven lighting or low contrast. By stretching the distribution of pixel values, the target features in the image are made more prominent, which helps the subsequent recognition algorithm work more effectively. Through these preprocessing steps, the original video image is optimized and becomes a preprocessed image that is more suitable for target and feature recognition.

[0083] In step A102, the object detection algorithm typically has strong feature learning and classification capabilities and is suitable for identifying objects with specific shapes and textures, such as lamps and vehicles. Applying the object detection algorithm to preprocessed images (e.g., noise reduction and contrast enhancement) can improve the detection rate and accuracy of lamps and vehicles, ensuring reliable acquisition of this critical information. This provides more accurate input for subsequent determination of lamp spacing based on lamp recognition results, as well as for determining vehicle position, calculating speed, and predicting arrival times based on vehicle recognition results. The object detection algorithm can use, but is not limited to, the YOLO algorithm.

[0084] In step A103, the Canny algorithm is a classic edge detection algorithm that is insensitive to noise and can detect clear edges in images. Lanes typically appear as lines with distinct edges in an image. The Canny algorithm effectively extracts this edge information, thereby identifying lanes. For example, two thresholds can be set to determine edge strength. Applying the Canny algorithm to an image that has undergone preprocessing (such as grayscale and noise reduction) can more accurately detect lane edges, providing precise baseline data for subsequent lane type determination and road space occupancy calculation based on lane recognition results. By employing different recognition algorithms for different identification targets, the overall recognition efficiency and effectiveness are improved.

[0085] Through the above-mentioned technical solutions, this application overcomes the problems of uneven illumination and noise interference encountered when directly using raw video images for recognition. Grayscale conversion processing reduces the data dimension and simplifies the processing process. Gaussian filtering effectively suppresses image noise and improves image quality. Histogram equalization processing enhances image contrast, making the target features in the image clearer. The target detection algorithm is used to identify lamps and vehicles, and the Canny algorithm is used to identify lanes. These algorithms are selected based on the characteristics of their respective recognition objects, improving recognition accuracy. This provides reliable recognition results for subsequent tunnel lighting control.

[0086] In some embodiments, step A2 comprises:

[0087] A201. Detect the identified lane using a Hough transform to determine whether the lane is a straight line.

[0088] A202. If no straight line is detected, the driving lane is determined to be a curve, and a curve fitting method is used to fit the curve equation of the driving lane.

[0089] In step A201, the identified lane is tested using a Hough transform to determine whether it is a straight line. The Hough transform is an image processing algorithm suitable for detecting straight lines in images (the detection process is known in the art and will not be described in detail here). By performing a Hough transform on the lane identified in step A1, it is possible to analyze whether it conforms to the mathematical characteristics of a straight line, thereby determining whether the lane is a straight line. This is the first step in determining the lane type.

[0090] Step A202 is a branch process based on the result of the previous step. If the Hough transform fails to detect a straight line in step A201, the lane is determined to be a curve. To enable subsequent accurate calculations based on the curve's characteristics, this step further utilizes a curve fitting method. Curve fitting mathematically derives a continuous curve equation from the identified discrete lane pixel points. For example, polynomial fitting or spline fitting can be used to fit the curve equation.

[0091] This application introduces the Hough transform to provide an objective method for determining whether a lane is a straight line. If it is a curve, its mathematical equation is obtained through curve fitting. This provides a clear technical basis for determining lane type. Accurately determining lane type allows subsequent calculations of lamp pixel spacing and road space occupancy based on lane type to use geometric models appropriate for straight or curved lines, improving the accuracy of these calculations.

[0092] In some embodiments, step A3 comprises:

[0093] A301. Divide the tunnel lengthwise into multiple detection zones based on preset lamp spacing.

[0094] A302. Extract the pixel coordinates of the lamp based on the lamp identification result;

[0095] A303. If the lane is a straight line, calculate the average linear pixel distance between adjacent lamps based on the lamp pixel coordinates and use this as the lamp pixel spacing.

[0096] A304. If the lane is curved, calculate the average pixel distance between adjacent lamps along the curve equation based on the lamp pixel coordinates and the lane curve equation as the lamp pixel spacing.

[0097] A305. Calculate the conversion factor between pixel distance and actual distance based on the luminaire pixel pitch and the luminaire spacing.

[0098] In step A301, the number of lamps included in each detection area can be preset, and the length of a single detection area can be determined according to the number and the preset lamp spacing, so that the detection area can be divided according to the length.

[0099] Step A302 extracts the pixel coordinates of the lamp. The lamp recognition result provides information about the lamp's location in the image. Extracting the lamp's pixel coordinates is the basis for subsequent distance calculations. For example, the center coordinates of the lamp's target detection frame can be used as the lamp's pixel coordinates, or the centroid coordinates of the lamp's region can be extracted.

[0100] Furthermore, steps A303 and A304 perform different pixel distance calculations according to the lane type determined in step A2. Step A2 determines whether the lane is a straight line or a curve, and fits a curve equation if it is a curve.

[0101] If step A2 determines that the lane is a straight line, step A303 is executed. In step A303, the linear distance between adjacent luminaires is calculated based on the extracted pixel coordinates. For example, the linear distance between pixel coordinates can be calculated using the Euclidean distance formula. To obtain a more accurate estimate of the conversion scale factor, the linear pixel distances between multiple pairs of adjacent luminaires can be calculated and averaged to determine the effective luminaire pixel spacing.

[0102] If step A2 determines that the lane is a curve, step A304 is executed. In step A304, the pixel distances between adjacent lamps along the curve equation are calculated using the extracted pixel coordinates of adjacent lamps and the lane curve equation fitted in step A2. For example, the pixel coordinates of adjacent lamps are projected onto the curve equation, and the arc length between the two projected points is calculated to obtain the pixel distances along the curve equation. Similarly, the average of the pixel distances along the curve between multiple pairs of adjacent lamps can be calculated as the effective lamp pixel spacing to improve the accuracy of the estimated lamp pixel spacing.

[0103] Therefore, different distance calculation methods are selected according to the lane type, avoiding the errors caused by simply using straight-line distance on curved sections, making the determination of lamp pixel spacing more accurate.

[0104] Finally, in step A305 , the preset lamp spacing is divided by the lamp pixel spacing calculated in step A303 or A304 to obtain a conversion ratio factor between the pixel distance and the actual distance.

[0105] Through the above-mentioned technical solution, this application solves the problem of inaccurate conversion scaling factors caused by simply calculating the straight-line distance between lamps, especially when the lane type is curved. By adopting different distance calculation methods based on the lane type, namely calculating the straight-line distance for straight lines and the distance along the curve for curved lanes, this solution can more accurately determine the actual pixel spacing between lamps, and thus more accurately estimate the conversion scaling factor. This improves the accuracy of the conversion scaling factor determination, provides more reliable support for subsequent steps such as vehicle position determination, speed calculation, and arrival time prediction, and improves the accuracy of these steps.

[0106] In some embodiments, step A4 comprises:

[0107] A401. Based on the vehicle identification results and the conversion scale factor, a target tracking algorithm is used to track the vehicle's position in each detection area and obtain the vehicle's movement trajectory.

[0108] A402. Calculate the vehicle's speed based on its trajectory;

[0109] A403. Calculate the arrival time of the vehicle at the next light fixture based on the vehicle's current position and the position of the next light fixture, combined with the vehicle's speed.

[0110] Among them, in step A401, the vehicle recognition result provides the initial pixel position information of the vehicle at a certain moment. The target tracking algorithm is applied to a continuous sequence of video frames to continuously determine the pixel position of the vehicle at different time points, and then convert it into a spatial position according to the conversion scale factor. By connecting the spatial positions of the vehicle at consecutive time points, the movement trajectory of the vehicle can be constructed. For example, a Kalman filter, optical flow method or deep learning tracker can be used to achieve continuous tracking of the vehicle. Obtaining the movement trajectory of the vehicle provides continuous historical data of the vehicle's movement, which provides a basis for subsequent speed calculation and arrival time prediction.

[0111] Furthermore, in step A402, after obtaining the vehicle's trajectory, the vehicle's speed can be calculated based on the trajectory data. For example, the instantaneous speed can be obtained by calculating the ratio of the vehicle's position change between two adjacent time points to the time interval, or the average speed can be obtained by averaging the position changes over a period of time. Calculating speed based on a continuous trajectory can reduce the impact of single-frame image recognition errors on speed calculation, thereby obtaining more accurate speed information.

[0112] In step A403, the vehicle's current position can be obtained from the latest data point in the movement trajectory. The position of the next lamp is predetermined or obtained through lamp identification. Combined with the vehicle speed calculated in step A402, basic kinematic principles can be used to predict the time required for the vehicle to reach the next lamp. For example, the distance between the vehicle's current position and the next lamp's position (if the driving lane is detected as a straight line in step A201, a straight line fitting method can also be used to fit the straight line equation of the driving lane; here, if the driving lane is a straight line, the distance refers to the actual distance corresponding to the straight-line distance between the vehicle's current position and the projection point of the next lamp's position on the straight line equation; if the driving lane is a curve, the distance refers to the actual distance corresponding to the arc length between the vehicle's current position and the projection point of the next lamp's position on the curve equation) divided by the vehicle's speed. Explicitly linking the vehicle's current position, target position (at the next lamp), and vehicle speed for calculation ensures the logic and accuracy of the arrival time prediction.

[0113] This technical solution improves the continuity and accuracy of vehicle position information by explicitly employing a target tracking algorithm to obtain the vehicle's continuous trajectory. Calculating vehicle speed based on this continuous trajectory more accurately reflects the vehicle's actual motion state and reduces speed calculation errors. Using the vehicle's current position, the next lamp position, and the calculated vehicle speed to predict arrival time provides reliable and accurate input for subsequent arrival-time-based lamp control, thereby improving the timeliness and accuracy of lamp control.

[0114] In some embodiments, step A5 comprises:

[0115] A501. Calculate the road area of ​​each detection area based on lane type and preset road width;

[0116] A502. Calculate the actual area of ​​the vehicle's target detection frame based on the pixel size of the vehicle's target detection frame and the conversion scale factor. Using the actual area of ​​the vehicle's target detection frame as the vehicle's occupied area, calculate the sum of the occupied areas of all vehicles in each detection area.

[0117] A503. Divide the total occupied area by the corresponding road area to obtain the road space occupancy rate of each detection area;

[0118] A504. Compare the road space occupancy rate with the preset occupancy rate threshold to determine the target color temperature and corresponding target power of the lamps in each detection area.

[0119] In step A501, the road area of ​​each detection area is calculated based on the lane type and the preset road width, providing an accurate benchmark for the subsequent calculation of road space occupancy. The lane type can be a straight line or a curve identified through image processing methods. The preset road width is a fixed value obtained based on actual tunnel design or measurement. When the lane is a straight line, the road area can be calculated based on the straight length and road width of the detection area. When the lane is a curve, the road area can be calculated based on the curve length and road width of the detection area, or other geometric calculation methods can be used. This allows for adaptability to tunnel sections of varying geometries.

[0120] In step A502, the vehicle's target detection frame is the rectangular area in the image that the vehicle recognition algorithm uses to mark the vehicle's position and size. The pixel width and pixel height of the target detection frame can be converted to actual width and height using a conversion scaling factor, thereby calculating the actual area. Using the actual area of ​​the target detection frame as the vehicle's footprint is a feasible method for quantifying the space occupied by vehicles on the road from video images. By summing the actual areas of the target detection frames of all identified vehicles within the same detection area, the total area occupied by vehicles on the road in that detection area at the current moment can be obtained.

[0121] In step A503, for each detection area, the total area occupied by vehicles is divided by the calculated road area to obtain the road space occupancy rate. The road space occupancy rate is a dimensionless ratio that quantitatively reflects the density of vehicles on the road within the detection area. A higher occupancy rate indicates traffic congestion, while a lower occupancy rate indicates sparse (or smooth) traffic.

[0122] In step A504, the calculated road space occupancy rate is compared with a preset occupancy threshold to determine the target lighting color temperature and corresponding target power. For example, when traffic density is high, a color temperature that helps increase driver alertness can be used, while when traffic density is low, a more comfortable color temperature can be used. In this way, the lighting color temperature can be dynamically adjusted based on actual traffic density.

[0123] This method reflects traffic density by quantitatively calculating road space occupancy, using this information to dynamically adjust lighting color temperature and power. This differs from traditional solutions that use fixed lighting or simply adjust brightness, enabling a more precise matching of lighting parameters to actual traffic conditions. By linking the calculated road space occupancy with pre-set rules (such as thresholds or interpolation), lighting parameters appropriate for the current traffic conditions are directly derived, providing key input for subsequent control of individual lamps' on / off, operating color temperature, and operating power. This dynamic adjustment based on actual traffic density helps improve energy efficiency while ensuring driving safety and visual comfort.

[0124] Preferably, step A504 may include:

[0125] If the road space occupancy rate is greater than a first occupancy rate threshold, determining the target color temperature as a first preset color temperature;

[0126] If the road space occupancy rate is less than the second occupancy rate threshold, the target color temperature is determined to be the second preset color temperature; if the first occupancy rate threshold is greater than the second occupancy rate threshold, the first preset color temperature is greater than the second preset color temperature;

[0127] If the road space occupancy rate is between the second occupancy rate threshold and the first occupancy rate threshold, determining the target color temperature based on a linear interpolation operation, or determining the target color temperature as a third preset color temperature; the third preset color temperature is greater than the second preset color temperature and less than the first preset color temperature;

[0128] Obtain the light efficiency parameters corresponding to the determined target color temperature and the luminous flux target value required by the tunnel lighting design;

[0129] Divide the luminous flux target value by the luminous efficiency parameter to obtain the target power corresponding to the target color temperature.

[0130] Among them, the calculated road space occupancy is used to determine the target color temperature and the corresponding target power. Specifically, the calculated road space occupancy is compared with the preset first occupancy threshold and the second occupancy threshold. For example, the first occupancy threshold can be set to 80%, and the second occupancy threshold can be set to 50%. If the road space occupancy is greater than the first occupancy threshold, it indicates that the traffic flow in the tunnel is large. At this time, the target color temperature is determined to be a higher first preset color temperature, such as 6500K. A higher color temperature helps to improve the driver's visual clarity, enhance visual acuity and alertness, thereby improving driving safety. If the road space occupancy is less than the second occupancy threshold, it indicates that the traffic flow in the tunnel is small. At this time, the target color temperature is determined to be a lower second preset color temperature, such as 5000K. A lower color temperature helps to create a more comfortable visual environment, thereby improving visual comfort.

[0131] If the road occupancy rate is between the second and first occupancy rate thresholds, meaning traffic flow is at a moderate level, two methods are available for determining the target color temperature. One method uses linear interpolation to determine the target color temperature. Based on the specific occupancy rate within this range, an interpolated target color temperature between the second and first preset color temperatures is calculated. For example, when the occupancy rate is S, a target color temperature between the second and first preset color temperatures can be calculated using the linear interpolation formula CT = CT2 + (CT1-CT2) * (S-S2) / (S1-S2), where CT is the target color temperature, CT1 is the first preset color temperature, CT2 is the second preset color temperature, S1 is the first occupancy rate threshold, and S2 is the second occupancy rate threshold. Another method directly determines the target color temperature as a third preset color temperature, which is set to be greater than the second preset color temperature and less than the first preset color temperature, for example, 5700K. Both methods enable the selection of a moderate color temperature under moderate traffic flow conditions.

[0132] After determining the target color temperature, the target power corresponding to that target color temperature needs to be determined. This is achieved by obtaining the luminous efficacy parameters of the lamp corresponding to the determined target color temperature and the target luminous flux value required by the tunnel lighting design. The luminous efficacy parameter reflects the efficiency of the lamp in converting electrical energy into light energy at a specific color temperature. It is usually measured in lumens per watt (lm / W) and can be obtained in advance through experiments or by consulting the lamp's technical manual. The luminous efficacy parameters of the same lamp may vary at different color temperatures. The target luminous flux value required by the tunnel lighting design is a preset, desired illumination level. For example, each lamp is required to provide a luminous flux of 10,000 lumens (10,000 lm). By dividing the target luminous flux value by the obtained luminous efficacy parameters corresponding to the target color temperature, the electrical power required to achieve that luminous flux, i.e., the target power, can be calculated. In mainstream LED lamps, increasing color temperature is often accompanied by an increase in luminous efficacy parameters. For example, if the target color temperature is 6500K, the corresponding luminous efficacy parameter is 120lm / W, and the target power is 10,000lm / 120lm / W, which is approximately 83.3W. If the target color temperature is 5000K, the corresponding luminous efficacy parameter is 100lm / W, and the target power is 10,000lm / 100lm / W, which is approximately 100W. This shows that high color temperatures can also reduce energy consumption. Therefore, the corresponding target power can be calculated for different target color temperatures, ensuring that the required light level is provided while adjusting the power output according to the color temperature difference.

[0133] By setting occupancy thresholds and preset color temperatures, and introducing linear interpolation or a third preset color temperature, this technical solution achieves refined grading or smooth adjustment of color temperature, enabling lighting parameter adjustments to better match actual traffic conditions. Furthermore, by obtaining the luminous efficiency parameters corresponding to the target color temperature and calculating the target power based on the target luminous flux value, this ensures that lighting requirements can be met at different color temperatures. Furthermore, the power is adjusted based on the differences in luminous efficiency at different color temperatures, improving energy efficiency.

[0134] In some embodiments, step A6 includes:

[0135] A601. Based on the speed and lamp spacing, determine the lamps in front of the vehicle as the target lamp group.

[0136] A602. Determine the on / off time of the target lighting group based on the arrival time;

[0137] A603. Turn on the target lamp group at the start time and set the operating color temperature of the target lamp group to the corresponding target color temperature and the operating power to the corresponding target power;

[0138] A604. If the vehicle is detected arriving at the target lamp group before the shutoff time, a delay time is calculated based on the vehicle's speed to delay shutoff of the target lamp group. If the vehicle does not arrive at the target lamp group before the shutoff time, the target lamp group is shut off.

[0139] Among them, in step A601, the vehicle speed can be multiplied by a preset reference time value to obtain an extended distance, and then the next lamp is used as the starting lamp, and based on the lamp spacing, the first lamp in front of the vehicle whose distance from the starting lamp is not less than the extended distance is determined as the end lamp, and all lamps from the starting lamp to the end lamp (including the starting lamp and the end lamp) are used as the target lamp group.

[0140] In step A602, the on-time (i.e., on-time) and initial off-time (i.e., initial off-time) for the target lamp group can be determined based on the predicted time of the vehicle's arrival at the next lamp. For example, the on-time can be set to a fixed time before the predicted vehicle arrival time to ensure that the lamps are already illuminated when the vehicle enters the lighting area. The initial off-time can be set to the predicted vehicle arrival time plus a preset delay. This achieves preliminary on-demand lighting control based on predictive information.

[0141] In step A603, when the determined on-time arrives, the target lamp group is activated. Simultaneously, the operating color temperature and power of the target lamp group are set to the calculated target color temperature and power. In this way, the lighting status of the lamps is not only correlated with the presence of vehicles, but also matches the actual traffic density and demand.

[0142] Furthermore, in step A604, a dynamic adjustment mechanism based on the vehicle's actual state is introduced to improve the accuracy and safety of the shutdown control. If the system confirms through detection that the vehicle has reached the area where the target lamp group is located before the initially determined shutdown time, this indicates that the vehicle has arrived as expected or slightly earlier. At this time, the system does not immediately extinguish the lamps according to the initial shutdown time. Instead, it calculates a delay time based on the vehicle's current actual speed. This delay time can be calculated based on the length of the target lamp group's lighting area and the vehicle's speed, ensuring that the lamps continue to illuminate until the vehicle safely leaves the area. For example, if the vehicle is moving faster, the calculated delay time may be shorter; if the speed is slower, the delay time will be extended accordingly. The target lamp group will not shut down until the calculated delay time is added to the initial shutdown time. This delayed shutdown strategy based on actual speed effectively avoids the problem of premature light extinguishing due to prediction errors or speed changes, ensuring driving safety.

[0143] On the other hand, if the system detects that a vehicle has not yet reached the area where the target lamp group is located when the initially determined off time arrives, this may mean that the vehicle speed is much lower than predicted, or the prediction is significantly off. In this case, to avoid unnecessary energy waste, the system will turn off the target lamp group according to the initially determined off time. This ensures that the target lamp group will not remain on for long periods when no vehicles require lighting.

[0144] By combining predicted arrival times for preliminary planning and making dynamic adjustments based on the actual arrival and speed of vehicles, especially through precise control of the timing of closing, this solution overcomes the limitations of relying solely on predictions, making tunnel lighting control more flexible, safer and energy-efficient.

[0145] Through the above-mentioned technical solution, this application solves the problems that may arise from relying solely on predicted vehicle arrival times to control lamp switching. By determining on- and off-times based on predicted arrival times, on-demand lighting is achieved, avoiding energy waste caused by lamps being on for extended periods before vehicles arrive. By introducing a delayed off-time mechanism based on the vehicle's real-time speed, the lamps ensure continuous illumination while the vehicle passes through the illuminated area, effectively preventing the impact of sudden light extinguishing on driving safety. By turning off the lamps when the preset off-time is reached but the vehicle has not arrived, energy use is further optimized. This solution improves the accuracy, flexibility, and safety of tunnel lighting control, allowing the lighting system to better adapt to actual traffic conditions.

[0146] It should be noted that in step A6, when there are multiple vehicles at the same time with the same lamp as one of the target lamps of the target lamp group, the opening time, closing time and delay time are calculated for each relevant vehicle according to the above steps. If there is a contradiction in the control state of the target lamp according to the calculation results, for example, according to the calculation results of vehicle A, the target lamp should be in the off state at time t, and according to the calculation results of vehicle B, the target lamp should be in the on state at time t, then the target lamp is kept in the on state at the corresponding time to ensure the driving safety of each vehicle.

[0147] Preferably, after step A603 and before step A604, the following steps may also be included:

[0148] A605. Use image detection methods to detect whether the lamps in the target lamp group are successfully turned on, and issue an abnormal warning if they are not successfully turned on.

[0149] Among them, after issuing the instruction to turn on the lamp and before deciding to turn off the lamp, a detection link for the actual on status of the lamp is added, which solves the problem that the lamp may not work normally according to the instruction. Specifically, through the image detection method, the video image is used to analyze whether the lamps in the target lamp group are emitting light, so as to determine whether they are successfully turned on. For example, after identifying the lamps in the target lamp group, the average pixel brightness within the target detection frame of the lamp is calculated and compared with the preset brightness threshold. If it is higher than the preset brightness threshold, it is determined that the lamp is successfully turned on; otherwise, it is determined that the lamp is not successfully turned on. This detection method based on visual feedback can directly verify the actual working status of the lamp. If it is found during the detection that at least one lamp fails to turn on successfully, the system will issue an abnormal warning so that the fault can be handled in time to ensure the reliability of tunnel lighting and driving safety.

[0150] Through the above technical solution, this application solves the problem that simply issuing an on command doesn't guarantee the lamps will actually turn on and function properly. By adding a video-based lamp status detection step, it's possible to directly verify whether the lamps are emitting light as instructed. This allows for timely detection of issues such as lamp failures or communication anomalies, which can be addressed by issuing an anomaly warning. This ensures the reliability of the tunnel lighting system, improves lighting effectiveness, and thus ensures driving safety.

[0151] refer to Figure 2 , the present application provides a tunnel lighting control system based on video surveillance, comprising a plurality of lamps 1 arranged at intervals along the tunnel, a number of cameras 2 and a control terminal 3;

[0152] Camera 2 is used to collect video images in the tunnel and send them to the control terminal 3;

[0153] Console 3 is used to execute:

[0154] Based on the video image, identify the lamp 1, the lane, and the vehicle (refer to step A1 above for the specific process);

[0155] Determine the lane type based on the lane recognition result; the lane type can be straight or curved (refer to step A2 above for the specific process);

[0156] Divide the tunnel into multiple detection zones based on the preset lamp spacing. Calculate the conversion factor between pixel distance and actual distance based on the lamp recognition results, lane type, and lamp spacing (see step A3 above for details).

[0157] Based on the vehicle identification results and the conversion scale factor, the target tracking algorithm is used to determine the vehicle's position in each detection area. This is used to calculate the vehicle's speed and predict the vehicle's arrival time at the next light fixture (see step A4 above for the specific process).

[0158] Calculate the road space occupancy rate of each detection area based on the vehicle recognition results, conversion scale factor, and lane type to determine the target color temperature and corresponding target power of the lamps in each detection area (refer to step A5 above for the specific process);

[0159] Based on the speed, lamp spacing, arrival time, target color temperature, and target power, the on / off, operating color temperature, and operating power of several lamps in front of the vehicle are controlled (for the specific process, refer to step A6 above).

[0160] In summary, this application has at least the following advantages:

[0161] 1. Multi-zone collaborative detection to avoid blind spots: Existing technologies often rely on a single detection method (such as radar or infrared sensors), which can easily lead to blind spots in certain areas and cause inaccurate lighting control. However, this invention integrates video detection and multi-zone speed measurement technology to achieve comprehensive monitoring of vehicle position and speed in tunnels, avoiding lighting system misjudgments caused by vehicle speed uncertainty in unmonitored areas and ensuring accurate and reliable lighting control.

[0162] 2. Dynamically adjust lamp on-time and color temperature: Existing technologies typically control lamps based on preset schedules or fixed patterns, which cannot adapt to changes in actual traffic flow and vehicle driving conditions. However, this invention uses real-time monitoring of vehicle speed and position, combined with target tracking technology, to dynamically adjust lamp on-time and off-time, as well as color temperature, ensuring effective lighting while reducing energy consumption, improving driving safety and energy efficiency.

[0163] 3. Intelligent anomaly detection and status monitoring: Existing systems have limited monitoring of lamp operating conditions, making it difficult to detect and address faults in a timely manner. This invention continuously analyzes the operating conditions of lamps through video data, enabling immediate detection of anomalies and issuing alarms, significantly improving system reliability and maintenance efficiency while reducing maintenance costs.

[0164] 4. Automatically generate detection areas and simplify deployment: Compared with existing technologies, this invention can automatically generate detection areas based on the spacing of tunnel lighting fixtures, eliminating the need for manual settings. This simplifies system deployment and debugging, improves system flexibility and adaptability, and reduces system deployment costs and time.

[0165] 5. Refined lighting control and optimized energy utilization: This invention achieves refined lighting control by calculating the road space occupancy rate within the detection area, adjusting the lighting intensity and color temperature as needed, further optimizing energy utilization, reducing operating costs, and providing a more comfortable visual environment for the driver.

[0166] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A tunnel lighting control method based on video monitoring, characterized in that: Including steps: A1. Capture video images inside the tunnel to identify lamps, lanes, and vehicles. A2. Determine the lane type based on the lane recognition results; the lane type includes straight or curved; A3. Divide the tunnel into multiple detection areas based on the preset lamp spacing and calculate the conversion factor between the pixel distance and the actual distance based on the lamp recognition results, lane type, and the lamp spacing; A4. Based on the vehicle identification results and the conversion scale factor, a target tracking algorithm is used to determine the vehicle's position in each detection area, calculate the vehicle's speed, and predict the vehicle's arrival time at the next light fixture. A5. Calculate the road space occupancy rate for each detection area based on the vehicle identification results, the conversion scaling factor, and the lane type to determine the target color temperature and corresponding target power of the lamps in each detection area; A6. Control the on / off, operating color temperature, and operating power of the plurality of lamps in front of the vehicle based on the speed, the lamp spacing, the arrival time, the target color temperature, and the target power.

2. The tunnel lighting control method based on video monitoring according to claim 1, characterized in that: Step A1 includes: A101 obtains the original video image in the tunnel, performs preprocessing, obtains the preprocessed video image; the preprocessing includes grayscale conversion processing for converting the color image into a grayscale image, Gaussian filtering processing and histogram equalization processing; A102. Use an object detection algorithm to identify lamps and vehicles in preprocessed video images. A103. Use the Canny algorithm to identify lanes in the preprocessed video image.

3. The tunnel lighting control method based on video monitoring according to claim 1, characterized in that: Step A2 includes: A201. Detect the identified lane using a Hough transform to determine whether the lane is a straight line. A202. If no straight line is detected, determine that the driving lane is a curve, and use a curve fitting method to fit the curve equation of the driving lane.

4. The tunnel lighting control method based on video monitoring according to claim 3 is characterized in that: Step A3 includes: A301. Divide the tunnel lengthwise into multiple detection zones based on preset lamp spacing. A302. Extract the pixel coordinates of the lamp based on the lamp identification result; A303. If the lane is a straight line, then the average straight line pixel distance between adjacent lamps is calculated based on the pixel coordinates of the lamps as the lamp pixel spacing; A304. If the lane is a curve, then based on the pixel coordinates of the lamps and the curve equation of the lane, calculate the average pixel distance between adjacent lamps along the curve equation as the lamp pixel spacing; A305. Calculate a conversion ratio factor between the pixel distance and the actual distance based on the lamp pixel pitch and the lamp pitch.

5. The tunnel lighting control method based on video monitoring according to claim 1, characterized in that: Step A4 includes: A401. Based on the vehicle identification results and the conversion scale factor, the target tracking algorithm is used to track the position of the vehicle in each detection area to obtain the vehicle's movement trajectory; A402. Calculate the vehicle speed based on the moving trajectory; A403. Calculate the arrival time of the vehicle at the next light fixture based on the vehicle's current position and the position of the next light fixture, combined with the vehicle's speed.

6. The tunnel lighting control method based on video surveillance according to claim 1, characterized in that: Step A5 includes: A501. Calculate the road area of ​​each detection area based on lane type and preset road width; A502. Calculate the actual area of ​​the vehicle's target detection frame based on the pixel size of the vehicle's target detection frame and the conversion scale factor. Using the actual area of ​​the vehicle's target detection frame as the vehicle's occupied area, calculate the sum of the occupied areas of all vehicles in each detection area. A503. Divide the total occupied area by the corresponding road area to obtain the road space occupancy rate of each detection area; A504. Compare the road space occupancy rate with the preset occupancy rate threshold to determine the target color temperature and corresponding target power of the lamps in each detection area.

7. The tunnel lighting control method based on video monitoring according to claim 6, characterized in that: Step A504 includes: If the road space occupancy rate is greater than a first occupancy rate threshold, determining the target color temperature as a first preset color temperature; If the road space occupancy rate is less than the second occupancy rate threshold, the target color temperature is determined to be the second preset color temperature; if the first occupancy rate threshold is greater than the second occupancy rate threshold, the first preset color temperature is greater than the second preset color temperature; If the road space occupancy rate is between the second occupancy rate threshold and the first occupancy rate threshold, determining the target color temperature based on a linear interpolation operation, or determining the target color temperature as a third preset color temperature; the third preset color temperature is greater than the second preset color temperature and less than the first preset color temperature; Obtain the light efficiency parameters corresponding to the determined target color temperature and the luminous flux target value required by the tunnel lighting design; The target power corresponding to the target color temperature is obtained by dividing the target luminous flux value by the light efficiency parameter.

8. The tunnel lighting control method based on video surveillance according to claim 1, characterized in that: Step A6 includes: A601. Based on the speed and the lamp spacing, determine several lamps in front of the vehicle as the target lamp group; A602. Determine the on and off time of the target lighting group based on the arrival time; A603. Turn on the target lamp group at the start time, and set the target lamp group's operating color temperature to the corresponding target color temperature, and its operating power to the corresponding target power; A604. If a vehicle is detected arriving at the target lamp group before the closing time, a delay time is calculated based on the vehicle's speed to delay closing the target lamp group; if the vehicle does not arrive at the target lamp group within the closing time, the target lamp group is closed.

9. The tunnel lighting control method based on video monitoring according to claim 8, characterized in that: After step A603 and before step A604, the following steps are also included: A605. Use an image detection method to detect whether the lamps in the target lamp group are successfully turned on, and issue an abnormal warning if they are not successfully turned on.

10. A tunnel lighting control system based on video surveillance, characterized in that: It includes multiple lamps set at intervals along the tunnel, several cameras and a control terminal; The camera is used to collect video images in the tunnel and send them to the control terminal; The control terminal is used to execute: Identify lamps, lanes, and vehicles based on the video image; Determine the lane type based on the lane recognition result; the lane type includes straight line or curve; Dividing the tunnel into multiple detection areas according to a preset lamp spacing, and calculating a conversion factor between pixel distance and actual distance based on the lamp recognition results, lane type, and the lamp spacing; Based on the vehicle identification result and the conversion scale factor, a target tracking algorithm is used to determine the position of the vehicle in each detection area, thereby calculating the vehicle's speed and predicting the vehicle's arrival time at the next light fixture; Calculating the road space occupancy rate of each detection area based on the vehicle recognition result, the conversion scale factor, and the lane type to determine the target color temperature and corresponding target power of the lamps in each detection area; According to the speed, the lamp spacing, the arrival time, the target color temperature, and the target power, the switching, operating color temperature, and operating power of the plurality of lamps in front of the vehicle are controlled.

Images