Safe passing method, device and equipment for invisible turning road section and storage medium

Abstract

The invention discloses a safe passing method, device and equipment for invisible turning road sections and a storage medium, and relates to the technical field of road traffic. According to the method, the characteristic that the longer the light wavelength is, the closer the color is to the warm color is utilized to define the flickering color of the light-emitting alarm part, and different flickering gears and color gears of the light-emitting alarm part are started by analyzing which preset gear the speed per hour of the vehicle closest to the turning position is located; the higher the flicker frequency of the light-emitting alarm piece is, the closer the flicker color to red, so that a stronger alarm impression is achieved, a driver can easily notice the reminding mechanism, the opposite road section is reminded by recognizing the current straight road section, and the function of bidirectional simultaneous reminding is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of road traffic, and particularly to a safe passing method, device, equipment and storage medium for a non-line-of-sight turning section. Background Art

[0002] During road design and construction, due to various factors such as land acquisition and demolition, or poor route alignment, the turning section may be non-line-of-sight, resulting in poor sight distance in the turning section and posing a significant potential safety hazard to passing vehicles and pedestrians. This situation is particularly prominent in low-grade roads, internal roads in communities, and indoor roads.

[0003] Normally, to reduce traffic accidents in the above situations, convex mirrors are generally installed in the above-mentioned sections with poor sight distance to remind oncoming vehicles of the upcoming meeting situation, so as to reduce safety risks.

[0004] However, the above measures have deficiencies. On the one hand, some sections, communities, and indoor areas are prohibited from honking, which conflicts with the requirement of honking before entering the turning intersection to remind oncoming passing vehicles. Secondly, due to reasons such as aging, damage of the convex mirror or inattentiveness of the driver, some drivers cannot master the situation of the oncoming lane through the convex mirror before entering the turning section, resulting in an increased probability of traffic accidents. Summary of the Invention

[0005] The main purpose of the present application is to provide a safe passing method, device, equipment and storage medium for a non-line-of-sight turning section to solve the problem that the safety of meeting vehicles in the turning section cannot be guaranteed in the prior art.

[0006] To achieve the above purpose, the present application provides the following technical solutions:

[0007] A safe passing method for a non-line-of-sight turning section, the non-line-of-sight turning section includes a turning point, and two straight sections respectively connected to both ends of the turning point. At least one side of the turning point has a visual obstruction that causes the two straight sections to be non-line-of-sight. A light-emitting alarm is installed on each surface of the visual obstruction facing the two straight sections respectively. The safe passing method includes:

[0008] Step S1, construct a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval;

[0009] Step S2, evenly divide the driving speed interval into several driving speed sub-intervals. Each driving speed sub-interval corresponds to a flashing frequency of a light-emitting alarm, and all flashing frequencies increase as the driving speed in the driving speed sub-interval increases;

[0010] Step S3: Obtain the vehicle counts for several preset time periods on the current straight section of the road.

[0011] Step S4: Take the maximum value among all the vehicle counts as the maximum load capacity of the current straight section of the road.

[0012] Step S5: Construct a load capacity interval with zero as the minimum value of the interval and the maximum load capacity as the maximum value of the interval.

[0013] Step S6: Evenly divide the load capacity interval into several load capacity sub - intervals. Each load capacity sub - interval corresponds to a flashing color of a light - emitting alarm device, and the wavelengths of all the flashing colors increase as the vehicle count in the load capacity sub - interval increases.

[0014] Step S7: Obtain the current driving speed of the vehicle closest to the turning point and the current load capacity of the current straight section of the road.

[0015] Step S8: Obtain the driving speed sub - interval that matches the current driving speed, the load capacity sub - interval that matches the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color.

[0016] Step S9: Send the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light - emitting alarm device on the opposite straight section of the road.

[0017] As a further improvement of this application, a shooting device is installed on each straight section of the road. After Step S9: Send the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light - emitting alarm device on the opposite straight section of the road, the following steps are included:

[0018] Step S10: Obtain the real - time image data of the current straight section of the road through the shooting device on the current straight section of the road.

[0019] Step S20: Use an object detection algorithm to determine whether there is a vehicle in the real - time image data. If not, execute Step S20; if so, execute Step S30.

[0020] Step S20: Turn off the light - emitting alarm device on the opposite straight section of the road.

[0021] Step S30: Determine whether the driving speed of the vehicle is zero. If so, execute Step S40.

[0022] Step S40: Determine that the vehicle is an illegally parked vehicle.

[0023] Step S50: Obtain the illegal parking image of the illegally parked vehicle through the shooting device on the current straight section of the road.

[0024] Step S60: Send the illegal parking image to an external monitoring terminal.

[0025] As a further improvement of the present application, in step S10, real-time image data of the current straight section is obtained through the shooting piece of the current straight section, and then, it includes:

[0026] In step S100, it is determined whether there are pedestrians in the real-time image data through the target detection algorithm. If there are, step S200 is executed;

[0027] In step S200, the opposite-side real-time image data of the opposite-side straight section is obtained through the shooting piece of the opposite-side straight section;

[0028] In step S300, it is determined whether there are vehicles in the opposite-side real-time image data through the target detection algorithm. If there are, step S400 is executed;

[0029] In step S400, the real-time distance between the pedestrian and the turning point is obtained;

[0030] In step S500, it is determined whether the real-time distance is less than or equal to a preset distance threshold. If so, step S600 is executed;

[0031] In step S600, the maximum flashing frequency and the maximum flashing color wavelength are generated and sent to the light-emitting alarm piece of the opposite-side straight section.

[0032] As a further improvement of the present application, in step S3, the numbers of vehicles in the current straight section based on several preset time periods are obtained, including:

[0033] In step S31, several image data of the current straight section are obtained through the shooting piece of the current straight section based on several preset time periods, and one image data is obtained for each preset time period;

[0034] In step S32, the current image data is evenly divided into several square grids;

[0035] In step S33, it is defined that the vehicle has the highest confidence, and several bounding boxes are predicted for all vehicles through all the square grids, and each bounding box includes at least one square grid;

[0036] In step S34, the confidence of each bounding box is obtained respectively;

[0037] In step S35, the bounding box with the maximum confidence is obtained and marked as the first-order bounding box;

[0038] In step S36, the intersection-over-union of the first-order bounding box and each other bounding box is calculated;

[0039] In step S37, all the bounding boxes with the intersection-over-union greater than or equal to the preset threshold are selected as the second-order bounding boxes;

[0040] Step S38: Obtain the second-order bounding box with the highest confidence level and define it as the detection box of a vehicle.

[0041] Step S39: Obtain the number of all detection boxes in the current image data, which is the number of vehicles on the current straight section based on the current preset time period.

[0042] As a further improvement of this application, in step S39, after obtaining the number of all detection boxes in the current image data, which is the number of vehicles on the current straight section based on the current preset time period, the following steps are included:

[0043] Step S1000: Integrate the number of vehicles in all preset time periods into a quantity data set.

[0044] Step S2000: Perform normalization processing on the quantity data set to obtain a normalized data set.

[0045] Step S3000: Divide the normalized data set into a training set and a validation set according to a preset ratio.

[0046] Step S4000: Define a neural network model with the input layer, hidden layer, and output layer connected in sequence for signal transmission.

[0047] Step S5000: Input the training set into the input layer and perform several trainings through the neural network model.

[0048] Step S6000: Based on each training, obtain the root mean square error between the validation set and the current training result respectively.

[0049] Step S7000: Obtain the minimum error among all root mean square errors, and obtain the training result corresponding to the minimum error as the vehicle quantity prediction model.

[0050] Step S8000: Predict several future vehicle quantities based on several preset prediction steps through the vehicle quantity prediction model.

[0051] Step S9000: Substitute the maximum value among all future vehicle quantities into step S5 to replace the maximum load capacity.

[0052] To achieve the above object, this application also provides the following technical solutions:

[0053] A safe passage device for a non-line-of-sight turning section, the safe passage device is applied to the above-mentioned safe passage method, and the safe passage device includes:

[0054] A driving speed interval construction module, which is used to construct a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval.

[0055] The driving speed interval division module is used to evenly divide the driving speed interval into several driving speed sub - intervals. Each driving speed sub - interval corresponds to a flashing frequency of a light - emitting alarm device, and all the flashing frequencies increase as the driving speed in the driving speed sub - intervals increases;

[0056] The vehicle quantity acquisition module is used to acquire several vehicle quantities in the current straight - ahead section based on several preset time periods;

[0057] The maximum load capacity definition module is used to take the maximum value of all vehicle quantities as the maximum load capacity of the current straight - ahead section;

[0058] The load capacity interval construction module is used to construct a load capacity interval with zero as the interval minimum value and the maximum load capacity as the interval maximum value;

[0059] The load capacity interval construction module is used to evenly divide the load capacity interval into several load capacity sub - intervals. Each load capacity sub - interval corresponds to a flashing color of a light - emitting alarm device, and the wavelengths of all the flashing colors increase as the vehicle quantity in the load capacity sub - intervals increases;

[0060] The driving vehicle parameter acquisition module is used to acquire the current driving speed of the driving vehicle closest to the turning point and the current load capacity of the current straight - ahead section;

[0061] The light - emitting alarm device parameter matching module is used to acquire the driving speed sub - interval that matches the current driving speed, the load capacity sub - interval that matches the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color;

[0062] The light - emitting alarm device parameter sending module is used to send the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light - emitting alarm device on the opposite straight - ahead section.

[0063] To achieve the above - mentioned purpose, the present application also provides the following technical solutions:

[0064] An electronic device includes a processor and a memory coupled to the processor. The memory stores program instructions executable by the processor; when the processor executes the program instructions stored in the memory, the above - mentioned safe passage method is implemented.

[0065] To achieve the above - mentioned purpose, the present application also provides the following technical solutions:

[0066] A storage medium stores program instructions, and when the program instructions are executed by a processor, the above - mentioned safe passage method can be implemented.

[0067] This application constructs a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; evenly divides the driving speed interval into several driving speed sub-intervals, each driving speed sub-interval corresponds to a flashing frequency of a light-emitting alarm device, and all flashing frequencies increase as the driving speed in the driving speed sub-interval increases; obtains the number of vehicles in a current straight section based on several preset time periods; takes the maximum value of all vehicle numbers as the maximum load capacity of the current straight section; constructs a load capacity interval with zero as the minimum value of the interval and the maximum load capacity as the maximum value of the interval; evenly divides the load capacity interval into several load capacity sub-intervals, each load capacity sub-interval corresponds to a flashing color of a light-emitting alarm device, and the wavelengths of all flashing colors increase as the number of vehicles in the load capacity sub-interval increases; obtains the current driving speed of the vehicle closest to the turning point and the current load capacity of the current straight section; obtains the driving speed sub-interval matching the current driving speed, the load capacity sub-interval matching the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color; sends the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light-emitting alarm device on the opposite straight section. This application utilizes the characteristic that the longer the wavelength of light, the closer the color is to warm colors to define the flashing color of the light-emitting alarm device, analyzes which preset gear the speed of the vehicle closest to the turning point is in to activate different flashing gears and color gears of the light-emitting alarm device, and when the speed is faster, the flashing frequency of the light-emitting alarm device is higher and the flashing color is closer to red, so as to achieve a stronger alarm visual effect, enabling the driver to more easily notice the above reminder mechanism, and this application reminds the opposite section by identifying the current straight section, realizing the function of two-way simultaneous reminder. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a schematic flowchart of the steps of an embodiment of the safe passing method for a non-through turning section of this application;

[0069] Figure 2 It is a schematic functional module diagram of an embodiment of the safe passing device for a non-through turning section of this application;

[0070] Figure 3 It is a schematic structural diagram of an embodiment of an electronic device of this application;

[0071] Figure 4 It is a schematic structural diagram of an embodiment of a storage medium of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying 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 the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0073] The terms "first", "second", and "third" in the present application are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", and "third" may explicitly or implicitly include at least one of such features. In the description of the present application, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present application are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes unlisted steps or units, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.

[0074] Referring to "embodiments" herein means that specific features, structures, or characteristics described in connection with the embodiments may be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0075] As Figure 1 shown, this embodiment provides an embodiment of a safe passing method for a non-through turning section. In this embodiment, the non-through turning section includes a turning point and two straight sections respectively connected to both ends of the turning point. At least one side of the turning point has a vision obstruction that causes the two straight sections to be non-through, and a luminous alarm is installed on each side of the vision obstruction facing the two straight sections.

[0076] Preferably, non-through turning sections are common in low-grade roads, internal roads in communities, and indoor roads; low-grade roads such as third-class roads, fourth-class roads and below, internal roads in communities are mostly roads between buildings, and indoor roads such as roads in underground garages.

[0077] Preferably, the safe passage method includes the following steps:

[0078] Step S1: Construct a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval.

[0079] Step S2: Evenly divide the driving speed interval into several driving speed sub - intervals. Each driving speed sub - interval corresponds to a flashing frequency of a light - emitting alarm device, and all the flashing frequencies increase as the driving speed in the driving speed sub - interval increases.

[0080] Preferably, the number of driving speed sub - intervals in the context can be set according to the flashing frequency gears of the light - emitting alarm device for the driving speed interval in this embodiment.

[0081] For example, if the flashing frequency gears of the light - emitting alarm device are three gears: slow flash, fast flash, and burst flash, corresponding to flashing once every 3 seconds, once every 1 second, and once every 0.25 seconds respectively, and according to the local speed limit of 30 km / h, the driving speed sub - intervals are set as [0 km / h, 10 km / h), [10 km / h, 20 km / h), [20 km / h, 30 km / h].

[0082] Step S3: Obtain the number of vehicles in a number of preset time periods for the current straight section of the road.

[0083] Preferably, the preset time period in this embodiment and the preset prediction steps in the following text can also be set as 3 minutes, 5 minutes, 10 minutes, etc.

[0084] Step S4: Take the maximum value of all the vehicle numbers as the maximum load capacity of the current straight section of the road.

[0085] For example, a certain road is a two - way single - lane road, the length of the straight section is 100 m. Calculate the lane with the driving direction towards the turning point. About 15 to 20 vehicles can be parked in 100 m.

[0086] It should be noted that the lane with the driving direction away from the turning point has completed passing each other and does not need to be calculated again.

[0087] Step S5: Construct a load capacity interval with zero as the minimum value of the interval and the maximum load capacity as the maximum value of the interval.

[0088] Step S6: Evenly divide the load capacity interval into several load capacity sub - intervals. Each load capacity sub - interval corresponds to a flashing color of a light - emitting alarm device, and the wavelengths of all the flashing colors increase as the number of vehicles in the load capacity sub - interval increases.

[0089] Preferably, the number of load capacity sub - intervals in the context can also be set according to the number of types of flashing colors of the light - emitting alarm device for the load capacity interval in this embodiment.

[0090] For example, the flashing colors of the light-emitting alarm device are green, yellow, and red, and the corresponding wavelengths are 500 nm to 577 nm, 577 nm to 597 nm, and 622 nm to 780 nm in sequence. The visible wavelengths increase in sequence. According to the above 100 m, the load capacity range can park about 15 to 20 vehicles. If 20 vehicles are used, the load quantum sub-ranges are [0 vehicles, 6 vehicles], [7 vehicles, 13 vehicles], and [14 vehicles, 20 vehicles] in sequence. The aforementioned vehicle quantity ranges use natural numbers.

[0091] Preferably, there are multiple types of flashing colors of the light-emitting alarm device. For example, if there are 7 types, they can be set to the following wavelengths:

[0092] Red: The wavelength range is 622 nm to 780 nm.

[0093] Orange: The wavelength range is 597 nm to 622 nm.

[0094] Yellow: The wavelength range is 577 nm to 597 nm.

[0095] Green: The wavelength range is 500 nm to 577 nm.

[0096] Cyan: The wavelength range is 470 nm to 500 nm.

[0097] Blue: The wavelength range is 430 nm to 470 nm.

[0098] Purple: The wavelength range is 380 nm to 430 nm.

[0099] As can be seen from the above, the number of corresponding load quantum sub-ranges is 7, and each load quantum sub-range has three natural numbers.

[0100] Preferably, in daily use, the three colors of red, yellow, and blue can be used to distinguish the misleading of the passable green.

[0101] It should be noted that during actual use, the three-color signal lights are sufficient and the seven colors are hardly used. For extremely sharp turning environments, the visible light wavelengths can be evenly divided into multiple parts, and then the color span performance of each part will be weakened.

[0102] Step S7, obtain the current driving speed of the driving vehicle closest to the turning point and the current load capacity of the current straight section.

[0103] Step S8, obtain the driving speed sub-range matching the current driving speed, the load quantum sub-range matching the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color.

[0104] For example, at this time, the current driving speed of the vehicle closest to the turning point has dropped to 15 km / h to prepare for turning, and there is only one vehicle in the current load of the current straight section. Then, the light-emitting alarm device on the opposite side flashes a green light at a frequency of once every three seconds.

[0105] Step S9: Send the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light-emitting alarm device on the opposite straight section.

[0106] Furthermore, each straight section is respectively equipped with a shooting device. After step S9: sending the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light-emitting alarm device on the opposite straight section, it includes:

[0107] Step S10: Obtain the real-time image data of the current straight section through the shooting device on the current straight section.

[0108] Step S20: Determine whether there is a vehicle in the real-time image data through the target detection algorithm. If not, execute step S20. If so, execute step S30.

[0109] Step S20: Turn off the light-emitting alarm device on the opposite straight section.

[0110] Step S30: Determine whether the driving speed of the vehicle is zero. If so, execute step S40.

[0111] Step S40: Determine that the vehicle is an illegally parked vehicle.

[0112] Step S50: Obtain the illegal parking image of the illegally parked vehicle through the shooting device on the current straight section.

[0113] Step S60: Send the illegal parking image to the external monitoring terminal.

[0114] Preferably, the design intention of steps S10 to S60 is to urge and prevent illegal parking behaviors at the turning point and ensure safe passage at the turning point.

[0115] Furthermore, after step S10: obtaining the real-time image data of the current straight section through the shooting device on the current straight section, it includes:

[0116] Step S100: Determine whether there are pedestrians in the real-time image data through the target detection algorithm. If so, execute step S200.

[0117] Step S200: Obtain the opposite real-time image data of the opposite straight section through the shooting device on the opposite straight section.

[0118] Step S300: Determine whether there is a vehicle in the opposite real-time image data through the target detection algorithm. If so, execute step S400.

[0119] Step S400, obtain the real-time distance between the pedestrian and the turning point.

[0120] Step S500, determine whether the real-time distance is less than or equal to a preset distance threshold. If so, execute Step S600.

[0121] Preferably, the preset distance threshold can be set to 5 meters, 10 meters, etc.

[0122] Step S600, generate the maximum flashing frequency and the maximum flashing color wavelength and send them to the light-emitting alarm device on the oncoming straight section.

[0123] Preferably, the maximum flashing frequency and the maximum flashing color wavelength can be set to a unique level to avoid confusion with the vehicle meeting function. For example, it flashes red every 0.1 second.

[0124] Preferably, the design intention of Steps S100 to S600 is to remind the vehicle to pay extra attention to oncoming pedestrians.

[0125] Further, in Step S3, obtain the number of vehicles in the current straight section based on a number of preset time periods.

[0126] Preferably, Step S3 and its sub-steps can be implemented by a target detection algorithm.

[0127] Preferably, the target detection algorithm can be implemented by the target detection algorithms VJ, HOG, DPMDetector; deep learning two-stage target detection algorithms RCNN, SPPNet, FastRCNN, FasterRCNN; target detection Trick algorithms FPN, CascadeRCNN; deep learning one-stage target detection algorithms Yolo, X, SSD, RetinaNet; deep learning Anchor-free target detection algorithms CornerNet, CenterNet, FCOS; target detection algorithms based on Transformer DETR, etc.

[0128] Specifically, Step S3 includes the following steps:

[0129] Step S31, based on a number of preset time periods, obtain a number of image data of the current straight section through the shooting device of the current straight section, and obtain one image data for each preset time period.

[0130] Step S32, evenly divide the current image data into a number of square grids.

[0131] Step S33, define that the vehicle has the highest confidence, and predict a number of bounding boxes for all vehicles through all square grids, and each bounding box includes at least one square grid.

[0132] Step S34: Obtain the confidence of each bounding box respectively.

[0133] Step S35: Obtain the bounding box with the highest confidence and mark it as the first-order bounding box.

[0134] Preferably, the intersection over union is the ratio obtained by dividing the intersection of the first-order bounding box and each bounding box by the union of the first-order bounding box and each bounding box (which can be the ratio of areas).

[0135] Step S36: Calculate the intersection over union of the first-order bounding box and each other bounding box respectively.

[0136] Step S37: Select all the bounding boxes with the intersection over union greater than or equal to the preset threshold as the second-order bounding boxes.

[0137] Step S38: Obtain the second-order bounding box with the highest confidence and define it as the detection box of a vehicle.

[0138] Step S39: Obtain the number of all detection boxes of the current image data, which is the number of vehicles on the current straight section based on the current preset time period.

[0139] Preferably, each grid is used to predict the coordinates, width, and height of N first-order bounding boxes, as well as the confidence of each first-order bounding box, that is, each grid needs to predict N×(4 + 1) values.

[0140] It can be understood that each grid needs to predict N (x, y, w, h, confidence); where (x, y) is the offset of the center of the first-order bounding box relative to the grid, (w, h) is the ratio of the first-order bounding box relative to the adjusted-size picture, and (confidence) is the confidence of the grid, with a value of 1 or 0.

[0141] Preferably, the confidence can be understood as whether there is a target in the current grid and the accuracy of the first-order bounding box.

[0142] For example: Suppose there is a target in an adjusted-size picture, and the width and height of the adjusted-size picture are (w a , h a ), then:

[0143] Divide the picture evenly into 7×7 (S×S) grids. If there is a grid at the center of the target, the coordinates of this grid are (x a , y a ). Suppose the coordinates of the center of the target are (x b , y b ), then the above offset can be calculated according to to obtain the above offset.

[0144] Preferably, in actual detection, if the predicted first-order bounding box and the actual bounding box perfectly overlap, the value of the intersection over union is 1. In the actual application process, the preset ratio is generally set to 0.5 first to determine whether the predicted second-order bounding box is correct, and the accuracy of the second-order bounding box is positively correlated with the intersection over union.

[0145] Preferably, the YOLO algorithm also needs to train the first-order bounding box to improve the accuracy of object detection.

[0146] Next, train the above training model through a preset target training set, and iteratively adjust the weights and biases of the training model a certain number of times through the backpropagation algorithm to reduce the value of the loss function of the training model.

[0147] Preferably, the loss function is

[0148] where is an indicator function indicating whether the j-th first-order bounding box in the i-th grid is responsible for the target, and the value is 1 or 0; x i , y i , w i , h i , C i correspond to the (x, y, w, h, confidence) prediction values of the i-th grid respectively.

[0149] It can be understood that the loss function includes the deviation of the coordinate values of the first-order bounding box, the deviation of the confidence, and the deviation of the prediction probability (or class deviation).

[0150] where is the midpoint loss of the first-order bounding box in the coordinate value deviation, is the width and height loss of the first-order bounding box in the coordinate value deviation, is the deviation of the confidence, is the deviation of the prediction probability (or class deviation).

[0151] where λ coord is the localization error penalty. Generally, λ coord = 5; S 2 is the above-mentioned S×S grids; B is the number of first-order bounding boxes; and are the estimated values of the abscissa and ordinate of the midpoint of the i-th first-order bounding box; and are the estimated values of the width and height of the i-th first-order bounding box; C i is the confidence of the i-th first-order bounding box; is the estimated value of the confidence of the i-th first-order bounding box; λ noobjis the confidence prediction loss. Generally, λ noobj = 0.5; p i (c) is the class probability of the i-th first-order bounding box; is the estimated value of the class probability of the i-th first-order bounding box; p i (c) corresponds to classes in .

[0152] It should be noted that since not every grid necessarily contains a target, if there is no target in the grid, it will cause the value of (confidence) to be 0, resulting in too large a gradient span in the subsequent backpropagation algorithm. Therefore, λ coord is introduced to control the loss of the predicted position of the first-order bounding box, and λ noobj is introduced to control the loss of no target in a single grid.

[0153] It should be noted that the above additional content is only for principle explanation, and the symbol meanings of the above additional content are not interoperable with the symbol meanings in other parts of this embodiment.

[0154] Further, in step S39, the number of all detection boxes of the current image data is the number of vehicles on the current straight section based on the current preset time period. After that, it includes:

[0155] Step S1000, integrating the number of vehicles in all preset time periods into a number data set.

[0156] Step S2000, performing normalization processing on the number data set to obtain a normalized data set.

[0157] Preferably, this embodiment preferably uses the normalization method of zero-mean normalization (Z-score standardization). This method standardizes the data based on the mean and standard deviation of the original data, and the processed data conforms to the standard normal distribution, that is, the mean is 0 and the standard deviation is 1. For the normalization method, this embodiment can also use batch normalization. Compared with simple normalization in previous neural network training, only the data of the input layer was normalized, but not the data in the middle layer. Although the data set of the input nodes was normalized, the data distribution after matrix multiplication of the input data was likely to change greatly, and as the number of network layers in the hidden layer deepened, the change in data distribution would become larger and larger. Therefore, batch normalization performs normalization processing in the middle layer of the neural network, making the training effect better.

[0158] Step S3000, dividing the normalized data set into a training set and a validation set according to a preset ratio.

[0159] Preferably, the preset ratio can be set to 8:2 to divide the normalized data set into a training set and a sample set in the ratio of 8:2.

[0160] Step S4000: Define a neural network model with signal connections between the input layer, the hidden layer, and the output layer in sequence.

[0161] Preferably, the neural network model is characterized by the following formula:

[0162]

[0163] where y is the neural network model; x n is the nth input node of the input layer, and each input node corresponds to a data in the training set. is the weight from the mth input node of the input layer to the nth input node of the hidden layer. is the bias connected to the nth input node of the hidden layer. is the bias of the output layer; tansig(·) is the activation function; the number in the parentheses of the symbol subscript is the layer number, the subscript (1) is the first layer, i.e., the input layer, and the subscript (1, 2) is from the first layer to the second layer, i.e., from the input layer to the hidden layer.

[0164] It should be noted that the above formula and formula symbols are only for principle explanation, and their meanings are not interoperable with those in other positions.

[0165] Step S5000: Input the training set into the input layer and perform several trainings through the neural network model.

[0166] Step S6000: Based on each training, obtain the root mean square error between the validation set and the training result of the current training respectively.

[0167] Step S7000: Obtain the minimum error among all the root mean square errors, and obtain the training result corresponding to the minimum error as the vehicle number prediction model.

[0168] Step S8000: Predict several future vehicle numbers based on several preset prediction steps through the vehicle number prediction model.

[0169] Step S9000: Substitute the maximum value among all the future vehicle numbers into the maximum load capacity replaced in Step S5.

[0170] Preferably, the design intention of Steps S1000 to S9000 is to realize the oncoming vehicle prediction function at the turning.

[0171] Preferably, training a neural network for a training model usually requires providing a large amount of data, i.e., a dataset; the dataset is generally divided into three categories, namely the above-mentioned training set, validation set, and test set.

[0172] Among them, one epoch is equal to the process of training once using all the samples in the training set. By training once, it means performing one forward pass and one back pass; when the number of samples in one epoch (i.e., the training set) is too large, training once may consume too much time, and it is not entirely necessary to use all the data in the training set every time. Then, the entire training set needs to be divided into multiple small pieces, that is, divided into multiple batches for training; one epoch consists of one or more batches. A batch is a part of the training set, and only a part of the data, that is, one batch, is used in each training process. The process of training one batch is one iteration.

[0173] Preferably, the neural network training specifically includes a perceptron. The perceptron consists of two layers of neurons. The input layer receives external input signals and then transmits them to the output layer. The output layer is an M-P neuron, and the step function is y j = f(∑ i w i ·x i -θ i ). Here, the step function is for principle explanation and does not communicate with other symbols.

[0174] Preferably, given the training dataset, the weights w i (i = 1, 2,..., n) and the training bias θ i can be learned. θ i can be understood as the weight w corresponding to a fixed value with fixed inputs of -1 and 0 i+1 .

[0175] Preferably, the number of neural network training times in this embodiment can be set to 10,000 times.

[0176] Preferably, the learning rate for the 1st to 5000th epochs can be set to 0.01, the learning rate for the 5001st to 7500th epochs can be set to 0.001, and the learning rate for the 7501st to 10000th epochs can be set to 0.0001.

[0177] It can be understood that the neural network training in this embodiment mainly includes the following ideas:

[0178] ① Initialize the weights and biases in the network.

[0179] Initialize the parameter values (the weights and biases of the output units, and the weights and biases of the hidden units are all parameters of the model). This is to activate the forward propagation, obtain the output values of the elements in each layer, and then obtain the value of the loss function.

[0180] ② Activate the forward propagation to obtain the output values of each layer and the expected values of the loss functions of each layer.

[0181] ③ Calculate the error terms of the output units and the error terms of the hidden units according to the loss function.

[0182] Calculate each error term, calculate the gradient of the parameter with respect to the loss function or calculate the partial derivative according to the chain rule of calculus. When taking the partial derivative of a vector or matrix in a composite function, the partial derivative of the inner function of the composite function always chooses left multiplication; when taking the partial derivative of a scalar in a composite function, the partial derivative of the inner function of the composite function can choose either left multiplication or right multiplication.

[0183] ④ Update the weights and biases in the neural network.

[0184] ⑤ Repeat steps ② to ④ until the loss function is less than the preset bias or the number of iterations is used up, and output the parameters at this time as the current optimal parameters.

[0185] In this embodiment, a driving speed interval is constructed with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; the driving speed interval is evenly divided into several driving speed sub-intervals, and each driving speed sub-interval corresponds to a flashing frequency of a light-emitting alarm device, and all flashing frequencies increase as the driving speed in the driving speed sub-interval increases; the number of vehicles in a current straight section is obtained based on several preset time periods; the maximum value of all vehicle numbers is used as the maximum load of the current straight section; a load interval is constructed with zero as the minimum value of the interval and the maximum load as the maximum value of the interval; the load interval is evenly divided into several load sub-intervals, and each load sub-interval corresponds to a flashing color of a light-emitting alarm device, and the wavelengths of all flashing colors increase as the number of vehicles in the load sub-interval increases; the current driving speed of the vehicle closest to the turning point and the current load of the current straight section are obtained; the driving speed sub-interval matching the current driving speed, the load sub-interval matching the current load, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color are obtained; the corresponding flashing frequency and the wavelength of the corresponding flashing color are sent to the light-emitting alarm device on the opposite straight section. This embodiment utilizes the characteristic that the longer the wavelength of light, the closer the color is to warm colors to define the flashing color of the light-emitting alarm device. By analyzing which preset gear the speed of the vehicle closest to the turning point is in, different flashing gears and color gears of the light-emitting alarm device are activated, and when the speed is faster, the flashing frequency of the light-emitting alarm device is higher and the flashing color is closer to red, so as to achieve a stronger alarm perception, enabling the driver to more easily notice the above reminder mechanism. Moreover, this embodiment reminds the opposite section by identifying the current straight section, realizing the function of two-way simultaneous reminder.

[0186] As Figure 2 shown, this embodiment provides an embodiment of a safety passing device for a non-through vision turning section. In this embodiment, the safety passing device is applied to the safety passing method in the above-mentioned embodiment.

[0187] Specifically, the safety passing device includes a driving speed interval construction module 1, a driving speed interval division module 2, a vehicle number acquisition module 3, a maximum load definition module 4, a load interval construction module 5, a load interval construction module 6, a driving vehicle parameter acquisition module 7, a light-emitting alarm device parameter matching module 8, and a light-emitting alarm device parameter sending module 9, which are electrically connected in sequence.

[0188] Among them, the driving speed interval construction module 1 is used to construct a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; the driving speed interval division module 2 is used to evenly divide the driving speed interval into several driving speed sub-intervals, and each driving speed sub-interval corresponds to a flashing frequency of a light-emitting alarm device, and all the flashing frequencies increase as the driving speed in the driving speed sub-interval increases; the vehicle quantity acquisition module 3 is used to acquire several vehicle quantities of the current straight section based on several preset time periods; the maximum load capacity definition module 4 is used to take the maximum value of all vehicle quantities as the maximum load capacity of the current straight section; the load capacity interval construction module 5 is used to construct a load capacity interval with zero as the minimum value of the interval and the maximum load capacity as the maximum value of the interval; the load capacity interval division module 6 is used to evenly divide the load capacity interval into several load capacity sub-intervals, and each load capacity sub-interval corresponds to a flashing color of a light-emitting alarm device, and the wavelengths of all the flashing colors increase as the vehicle quantity in the load capacity sub-interval increases; the driving vehicle parameter acquisition module 7 is used to acquire the current driving speed of the driving vehicle closest to the turning point and the current load capacity of the current straight section; the light-emitting alarm device parameter matching module 8 is used to acquire the driving speed sub-interval matched with the current driving speed, the load capacity sub-interval matched with the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color; the light-emitting alarm device parameter sending module 9 is used to send the corresponding flashing frequency and the wavelength of the corresponding flashing color to the light-emitting alarm device on the opposite straight section.

[0189] Furthermore, the safe passage device further includes a straight section real-time image data acquisition module, a real-time image data vehicle detection module, an opposite straight section light-emitting alarm device closing module, a vehicle driving speed judgment module, an illegally parked vehicle determination module, an illegally parked image acquisition module, and an illegally parked image sending module that are electrically connected in sequence; the straight section real-time image data acquisition module is electrically connected to the light-emitting alarm device parameter sending module 9.

[0190] Among them, the straight section real-time image data acquisition module is used to acquire the real-time image data of the current straight section through the shooting device of the current straight section; the real-time image data vehicle detection module is used to judge whether there is a vehicle in the real-time image data through a target detection algorithm; the opposite straight section light-emitting alarm device closing module is used to close the light-emitting alarm device on the opposite straight section if there is no vehicle; the vehicle driving speed judgment module is used to judge whether the driving speed of the vehicle is zero if there is a vehicle; the illegally parked vehicle determination module is used to determine that the vehicle is an illegally parked vehicle if the speed is zero; the illegally parked image acquisition module is used to acquire the illegally parked image of the illegally parked vehicle through the shooting device of the current straight section; the illegally parked image sending module is used to send the illegally parked image to an external monitoring terminal.

[0191] Further, the safe passage device further includes a real-time image data pedestrian detection module, a oncoming real-time image data acquisition module, a oncoming real-time image data vehicle detection module, a real-time distance acquisition module for the pedestrian and the turning point, a real-time distance judgment module for the pedestrian and the turning point, and a maximum state generation and sending module for the light-emitting alarm device, which are electrically connected in sequence; the real-time image data pedestrian detection module is electrically connected to the real-time image data acquisition module of the straight section.

[0192] Among them, the real-time image data pedestrian detection module is used to judge whether there are pedestrians in the real-time image data through the target detection algorithm; the oncoming real-time image data acquisition module is used to, if there are pedestrians, acquire the oncoming real-time image data of the oncoming straight section through the shooting device of the oncoming straight section; the oncoming real-time image data vehicle detection module is used to judge whether there are vehicles in the oncoming real-time image data through the target detection algorithm; the real-time distance acquisition module for the pedestrian and the turning point is used to, if there are pedestrians, acquire the real-time distance between the pedestrian and the turning point; the real-time distance judgment module for the pedestrian and the turning point is used to judge whether the real-time distance is less than or equal to the preset distance threshold; the maximum state generation and sending module for the light-emitting alarm device is used to, if so, generate the maximum flashing frequency, the maximum flashing color wavelength and send them to the light-emitting alarm device of the oncoming straight section.

[0193] Further, the vehicle quantity acquisition module 3 specifically includes a first vehicle quantity acquisition sub-module, a second vehicle quantity acquisition sub-module, a third vehicle quantity acquisition sub-module, a fourth vehicle quantity acquisition sub-module, a fifth vehicle quantity acquisition sub-module, a sixth vehicle quantity acquisition sub-module, a seventh vehicle quantity acquisition sub-module, an eighth vehicle quantity acquisition sub-module, and a ninth vehicle quantity acquisition sub-module, which are electrically connected in sequence; the first vehicle quantity acquisition sub-module is electrically connected to the driving speed interval division module 2, and the ninth vehicle quantity acquisition sub-module is electrically connected to the maximum load definition module 4.

[0194] Among them, the first vehicle quantity acquisition sub-module is used to obtain a plurality of image data of the current straight road section through the shooting components of the current straight road section based on a plurality of preset time periods, and one image data is obtained for each preset time period; the second vehicle quantity acquisition sub-module is used to evenly divide the current image data into a plurality of square grids; the third vehicle quantity acquisition sub-module is used to define that a vehicle has the highest confidence, and predict a plurality of bounding boxes for all vehicles through all the square grids, and each bounding box includes at least one square grid; the fourth vehicle quantity acquisition sub-module is used to respectively obtain the confidence of each bounding box; the fifth vehicle quantity acquisition sub-module is used to obtain the bounding box with the largest confidence and mark it as the first-order bounding box; the sixth vehicle quantity acquisition sub-module is used to calculate the intersection-over-union ratio of the first-order bounding box with each other bounding box respectively; the seventh vehicle quantity acquisition sub-module is used to select all the bounding boxes with the intersection-over-union ratio greater than or equal to the preset threshold as the second-order bounding boxes; the eighth vehicle quantity acquisition sub-module is used to obtain the second-order bounding box with the highest confidence and define it as the detection box of a vehicle; the ninth vehicle quantity acquisition sub-module is used to obtain the number of all detection boxes of the current image data, which is the vehicle quantity of the current straight road section based on the current preset time period.

[0195] Further, the safe passage device further includes a vehicle quantity integration module, a quantity data set normalization processing module, a normalized data set division module, a neural network model definition module, a neural network model training module, a root mean square error acquisition module, a vehicle quantity prediction model acquisition module, a future vehicle quantity prediction module, and a future vehicle quantity substitution module that are electrically connected in sequence; the vehicle quantity integration module is electrically connected to the ninth vehicle quantity acquisition sub-module.

[0196] Among them, the vehicle quantity integration module is used to integrate the vehicle quantities of all preset time periods into a quantity data set; the quantity data set normalization processing module is used to perform normalization processing on the quantity data set to obtain a normalized data set; the normalized data set division module is used to divide the normalized data set into a training set and a validation set according to a preset ratio; the neural network model definition module is used to define a neural network model in which the input layer, the hidden layer, and the output layer are sequentially connected by signals; the neural network model training module is used to input the training set into the input layer and perform several trainings through the neural network model; the root mean square error acquisition module is used to respectively obtain the root mean square error between the validation set and the current training result based on each training; the vehicle quantity prediction model acquisition module is used to obtain the minimum error among all the root mean square errors, and obtain the training result corresponding to the minimum error as the vehicle quantity prediction model; the future vehicle quantity prediction module is used to predict a plurality of future vehicle quantities based on a plurality of preset prediction steps through the vehicle quantity prediction model; the future vehicle quantity substitution module is used to substitute the maximum value among all the future vehicle quantities into the load capacity interval construction module 5 to replace the maximum load capacity.

[0197] It should be noted that this embodiment is a functional module embodiment based on the above method embodiment. For the preferred, extended, limited, exemplified, and principle description parts of this embodiment, please refer to the above embodiment, and this embodiment will not be elaborated here.

[0198] In this embodiment, a driving speed interval is constructed with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; the driving speed interval is evenly divided into several driving speed sub-intervals, and each driving speed sub-interval corresponds to a flashing frequency of a light-emitting alarm device, and all flashing frequencies increase as the driving speed in the driving speed sub-interval increases; the number of vehicles in a current straight section based on several preset time periods is obtained; the maximum value of all vehicle numbers is used as the maximum load capacity of the current straight section; a load capacity interval is constructed with zero as the minimum value of the interval and the maximum load capacity as the maximum value of the interval; the load capacity interval is evenly divided into several load capacity sub-intervals, and each load capacity sub-interval corresponds to a flashing color of a light-emitting alarm device, and the wavelengths of all flashing colors increase as the number of vehicles in the load capacity sub-interval increases; the current driving speed of the vehicle closest to the turning point and the current load capacity of the current straight section are obtained; the driving speed sub-interval matching the current driving speed, the load capacity sub-interval matching the current load capacity, as well as the corresponding flashing frequency and the wavelength of the corresponding flashing color are obtained; the corresponding flashing frequency and the wavelength of the corresponding flashing color are sent to the light-emitting alarm device on the opposite straight section. This embodiment utilizes the characteristic that the longer the wavelength of light, the closer the color is to warm colors to define the flashing color of the light-emitting alarm device, analyzes which preset gear the speed of the vehicle closest to the turning point is in to turn on different flashing gears and color gears of the light-emitting alarm device, and when the speed is faster, the flashing frequency of the light-emitting alarm device is higher and the flashing color is closer to red, so as to achieve a stronger alarm visual effect, enabling the driver to more easily notice the above reminder mechanism, and this embodiment reminds the opposite section by identifying the current straight section, realizing the function of two-way simultaneous reminder.

[0199] Figure 3 An embodiment of the electronic device of the present application is shown. Refer to Figure 3 , the electronic device 10 includes a processor 101 and a memory 102 coupled to the processor 101.

[0200] The memory 102 stores program instructions for implementing the safe passage method for non-line-of-sight turning sections in any of the above embodiments.

[0201] The processor 101 is used to execute the program instructions stored in the memory 102 to perform safe passage for non-line-of-sight turning sections.

[0202] Among them, the processor 101 can also be referred to as a CPU (Central Processing Unit). The processor 101 may be an integrated circuit chip with signal processing capabilities. The processor 101 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0203] Furthermore, Figure 4 is a schematic structural diagram of a storage medium according to an embodiment of the present application. Refer to Figure 4 , the storage medium 11 of the embodiment of the present application stores program instructions 111 that can implement all the above methods. Among them, the program instructions 111 can be stored in the above storage medium in the form of a software product, including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media that can store program codes such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, or terminal devices such as computers, servers, mobile phones, and tablets.

[0204] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling, direct coupling, or communication connection can be through some interfaces, and the indirect coupling or communication connection of devices or units can be in an electrical, mechanical, or other form.

[0205] In addition, in each embodiment of the present application, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. The above is only the implementation manner of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be included in the patent protection scope of the present application by the same token.

[0206] The specific implementation manners of the present application have been described in detail above, but they are only examples, and the present application is not limited to the specific implementation manners described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application. Therefore, all equivalent transformations, modifications, improvements, etc. made without departing from the spirit and principle of the present application should be covered by the scope of the present application.

Claims

1. A method for safe passage on a blind turning section, wherein the blind turning section comprises a turning point and two straight sections connected to both ends of the turning point, at least one side of the turning point has a visual field obstructing the two straight sections, and a luminous alarm is installed on each side of the visual field obstructing the two straight sections, characterized in that: The safe passage method includes: Step S1, constructing a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; Step S2, dividing the driving speed interval into a plurality of driving speed sub-intervals, each driving speed sub-interval corresponds to a flashing frequency of the light emitting alarm component, and all flashing frequencies increase as the driving speed of the driving speed sub-interval increases; Step S3, obtaining the number of vehicles on the current straight road section based on a number of preset time periods; Step S4, taking the maximum value of all vehicle numbers as the maximum load of the current straight road section; Step S5, constructing a load interval with zero as the minimum value of the interval and the maximum load as the maximum value of the interval; Step S6, dividing the load volume interval into a plurality of load quantum intervals, each load quantum interval corresponds to a flashing color of the light-emitting alarm component, and the wavelength of all flashing colors increases as the number of vehicles in the load quantum interval increases; Step S7, obtaining the current speed of the vehicle closest to the turning point and the current load of the current straight section; Step S8, obtaining a driving speed sub-interval matching the current driving speed, a load sub-interval matching the current load, and a corresponding flashing frequency and a corresponding flashing color wavelength; Step S9, sending the corresponding flashing frequency and the corresponding flashing color wavelength to the luminous alarm component of the opposite straight road section.

2. The safe passage method according to claim 1, wherein each straight section is provided with a camera, characterized in that: Step S9, sending the corresponding flashing frequency and the corresponding flashing color wavelength to the luminous alarm component of the opposite straight road section, and then comprising: Step S10, obtaining real-time image data of the current straight road section through a camera of the current straight road section; Step S20, determining whether there is a vehicle in the real-time image data by a target detection algorithm, if not, executing step S20, if yes, executing step S30; Step S20, turning off the luminous alarm device of the opposite straight road section; Step S30, determining whether the vehicle's travel speed is zero, if it is zero, executing step S40; Step S40, determining that the vehicle is an illegally parked vehicle; Step S50, obtaining an illegally parked image of the illegally parked vehicle by photographing the current straight road section; Step S60: sending the illegal parking image to an external monitoring terminal.

3. The safe passage method according to claim 2, characterized in that: Step S10, obtaining real-time image data of the current straight road section through a camera of the current straight road section, and then comprising: Step S100, determining whether there is a pedestrian in the real-time image data by using the target detection algorithm, and if so, executing step S200; Step S200, obtaining real-time image data of the opposite side of the straight road section through a camera of the opposite side of the straight road section; Step S300, determining whether there is a vehicle in the opposite side real-time image data by the target detection algorithm, and if so, executing step S400; Step S400, obtaining the real-time distance between the pedestrian and the turning point; Step S500, determining whether the real-time distance is less than or equal to a preset distance threshold, if so, executing step S600; Step S600, generating a maximum flashing frequency and a maximum flashing color wavelength and sending them to a light emitting alarm component on the opposite straight road section.

4. The safe passage method according to claim 2, characterized in that: Step S3, obtaining the number of vehicles on the current straight road section based on a number of preset time periods, including: Step S31, acquiring a plurality of image data of the current straight road section through a camera of the current straight road section based on a plurality of preset time periods, wherein one image data is acquired in each preset time period; Step S32, dividing the current image data into a plurality of square grids on average; Step S33, defining the vehicle with the highest confidence, and predicting a number of bounding boxes for all vehicles through all square grids, each bounding box including at least one square grid; Step S34, obtaining the confidence of each bounding box respectively; Step S35, obtaining a bounding box with the highest confidence and marking it as a first-order bounding box; Step S36, calculating the intersection-over-union ratio between the first-order bounding box and each of the other bounding boxes; Step S37, selecting all bounding boxes whose intersection-over-union ratio is greater than or equal to a preset threshold as second-order bounding boxes; Step S38, obtaining a second-order bounding box with the highest confidence and defining it as a detection box of a vehicle; Step S39, obtaining the number of all detection frames of the current image data is the number of vehicles in the current straight section based on the current preset time period.

5. The safe passage method according to claim 4, characterized in that: Step S39, the number of all detection frames of the current image data is obtained, which is the number of vehicles on the current straight section based on the current preset time period. include: Step S1000, integrating the number of vehicles in all preset time periods into a quantity data set; Step S2000, normalizing the quantity data set to obtain a normalized data set; Step S3000, dividing the normalized data set into a training set and a validation set according to a preset ratio; Step S4000, defining a neural network model in which an input layer, a hidden layer, and an output layer are sequentially connected; Step S5000, inputting the training set into the input layer, and performing several trainings through the neural network model; Step S6000, obtaining a root mean square error between the validation set and the current training result based on each training; Step S7000, obtaining a minimum error value among all root mean square errors, and obtaining a training result corresponding to the minimum error value as a vehicle quantity prediction model; Step S8000, predicting a number of future vehicle numbers based on a number of preset prediction steps using the vehicle number prediction model; Step S9000, substituting the maximum value of all future vehicle quantities into step S5 to replace the maximum load.

6. A safe passage device for a blind turning section, the safe passage device being applied to a safe passage method as claimed in any one of claims 1 to 5, characterized in that: The safe passage device comprises: A driving speed interval construction module is used to construct a driving speed interval with zero as the minimum value of the interval and the speed limit of the current road as the maximum value of the interval; A driving speed interval division module, used for evenly dividing the driving speed interval into a plurality of driving speed sub-intervals, each driving speed sub-interval corresponds to a flashing frequency of the light-emitting alarm component, and all flashing frequencies increase as the driving speed of the driving speed sub-interval increases; A vehicle quantity acquisition module is used to acquire the number of vehicles on the current straight road section based on a number of preset time periods; The maximum load definition module is used to take the maximum value of all vehicle numbers as the maximum load of the current straight section; A load interval construction module, used to construct a load interval with zero as the minimum value of the interval and the maximum load as the maximum value of the interval; A load volume interval construction module, used for evenly dividing the load volume interval into a plurality of load quantum intervals, each load quantum interval corresponds to a flashing color of a light-emitting alarm component, and the wavelength of all flashing colors increases as the number of vehicles in the load quantum interval increases; A vehicle parameter acquisition module is used to acquire the current speed of the vehicle closest to the turning point and the current load of the current straight section; The light-emitting alarm parameter matching module is used to obtain the driving speed sub-interval matching the current driving speed, the load sub-interval matching the current load, and the corresponding flashing frequency and the corresponding flashing color wavelength; The luminous alarm component parameter sending module is used to send the corresponding flashing frequency and the corresponding flashing color wavelength to the luminous alarm component on the opposite straight section of the road.

7. An electronic device, characterized in that: It comprises a processor and a memory coupled to the processor, wherein the memory stores program instructions executable by the processor; when the processor executes the program instructions stored in the memory, the safe passage method as described in any one of claims 1 to 5 is implemented.

8. A storage medium, characterized in that: The storage medium stores program instructions, which, when executed by a processor, can implement the safe passage method according to any one of claims 1 to 5.

Images