A system and method for preventing rear-end collision at zebra crossings without signal lights

Abstract

The present invention relates to the field of traffic safety, and more specifically to a system and method for preventing rear-end collisions at zebra crossings without traffic lights. The system uses a braking prediction neural network model to predict whether a vehicle will brake and the magnitude of its deceleration when crossing a zebra crossing without traffic lights. The system then preemptively illuminates the brake lights based on the distance between the vehicle and the following vehicle, providing an early warning to the following vehicle and reducing the risk of traffic accidents.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety, and in particular to a system and method for preventing rear-end collisions at zebra crossings without signal lights. Background Art

[0002] The function of zebra crossings is to guide pedestrians to cross the road safely. Generally, vehicles should give way to pedestrians at zebra crossings without traffic lights. The purpose is to fully ensure the safety of pedestrians and avoid traffic accidents.

[0003] When a vehicle is driving and the driver is not paying attention (for example, the driver is looking down at the mobile phone) and is about to cross a zebra crossing without traffic lights, the driver fails to notice the pedestrians who are about to cross the road early and slow down in advance. The driver only suddenly realizes it when he is closer to the pedestrians, and then slams on the brakes to reduce the speed, and stops in a very short time to avoid the pedestrians, causing the driver of the vehicle behind to have no time to react and a rear-end collision to occur. Summary of the Invention

[0004] In view of the problems existing in the prior art, the object of the present invention is to provide a system and method for preventing rear-end collisions at zebra crossings without signal lights.

[0005] In order to achieve the above objectives, the present invention adopts the following technical solutions to achieve them.

[0006] A rear-end collision prevention system at a zebra crossing without a signal light, comprising a driver distraction sensing module, a pedestrian sensing module, a first millimeter-wave radar, a second millimeter-wave radar, and a processor;

[0007] The driver distraction sensing module includes an in-vehicle camera and a first image processor; the in-vehicle camera is used to capture the driver's facial image; the first image processor is used to determine whether the driver is distracted based on the driver's facial image;

[0008] The external pedestrian sensing module includes an external camera and a second image processor; the external camera is used to capture images of the road environment in front of the vehicle; the second image processor is used to determine whether a pedestrian is about to cross the zebra crossing or is crossing the zebra crossing based on the road environment image;

[0009] The first millimeter-wave radar is used to collect the distance between the vehicle and pedestrians about to cross the zebra crossing or currently crossing the zebra crossing, as well as the pedestrians' movement direction and speed;

[0010] The second millimeter-wave radar is used to collect the distance between the vehicle and the following vehicle;

[0011] The processor is used to predict whether the vehicle should brake and the amount of deceleration during braking, and is also used to control the vehicle's brake lights to light up.

[0012] A method for preventing rear-end collisions at a zebra crossing without a signal light, based on the above-mentioned system for preventing rear-end collisions at a zebra crossing without a signal light, comprises the following steps:

[0013] Step 1: Establish and train a braking prediction neural network model to obtain a trained braking prediction neural network model;

[0014] Step 2: While the vehicle is driving, the in-vehicle camera captures an image of the driver's face. The first image processor uses a machine learning model to determine whether the driver is distracted based on the facial image. The exterior camera captures an image of the road environment in front of the vehicle. The second image processor uses a machine vision method to identify the positions of pedestrians and zebra crossings based on the road environment image and determines whether a pedestrian is about to cross or is currently crossing the zebra crossing based on the pedestrian's movement direction and speed.

[0015] Step 3: When a pedestrian is about to cross the zebra crossing or is crossing the zebra crossing and the driver is distracted, and the distance between the vehicle and the pedestrian is less than or equal to 30m, record the vehicle speed v0 at the moment when the driver suddenly sees the pedestrian and is no longer distracted;

[0016] Step 4: The braking prediction neural network model predicts whether the vehicle will brake and the deceleration during braking based on the vehicle speed v0 at the moment when the driver suddenly sees the pedestrian and is no longer distracted, the pedestrian's moving speed v1, the distance d1 between the vehicle and the pedestrian, and the relative angle θ between the vehicle and the pedestrian.

[0017] Step 5: When it is predicted that the vehicle is about to brake, the processor controls the brake lights to light up;

[0018] When the distance between the vehicle and the following vehicle is less than 20m, and the vehicle is predicted to move at a speed of 3.5m / s 2 to 5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake lights will turn on immediately.

[0019] If the distance between the vehicle and the following vehicle is between 20m and 30m, and the vehicle is predicted to move at a speed of 2.5m / s 2 to 3.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.25s.

[0020] If the distance between the vehicle and the following vehicle is greater than 30m, and the vehicle is predicted to move at 0m / s 2 to 2.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.5s.

[0021] Compared with the existing technology, the beneficial effects of the present invention are: using a braking prediction neural network model to predict whether the vehicle will brake and the deceleration during braking, lighting the brake lights in advance according to the distance between the vehicle and the following vehicle, and giving advance warnings to the following vehicle, thereby reducing the risk of traffic accidents. DETAILED DESCRIPTION

[0022] The embodiments of the present invention will be described in detail below with reference to examples. However, those skilled in the art will understand that the following examples are only used to illustrate the present invention and should not be construed as limiting the scope of the present invention.

[0023] A rear-end collision prevention system at a zebra crossing without a signal light, comprising a driver distraction sensing module, a pedestrian sensing module, a first millimeter-wave radar, a second millimeter-wave radar, and a processor;

[0024] The driver distraction sensing module includes an in-vehicle camera and a first image processor; the in-vehicle camera is used to capture the driver's facial image; the first image processor is used to determine whether the driver is distracted based on the driver's facial image;

[0025] The external pedestrian sensing module includes an external camera and a second image processor; the external camera is used to capture images of the road environment in front of the vehicle; the second image processor is used to determine whether a pedestrian is about to cross the zebra crossing or is crossing the zebra crossing based on the road environment image;

[0026] The first millimeter-wave radar is used to collect the distance between the vehicle and pedestrians about to cross the zebra crossing or currently crossing the zebra crossing, as well as the pedestrians' movement direction and speed;

[0027] The second millimeter-wave radar is used to collect the distance between the vehicle and the following vehicle;

[0028] The processor is used to predict whether the vehicle should brake and the amount of deceleration during braking, and is also used to control the vehicle's brake lights to light up.

[0029] A method for preventing rear-end collisions at a zebra crossing without a signal light comprises the following steps:

[0030] Step 1: Establish and train a braking prediction neural network model to obtain a trained braking prediction neural network model;

[0031] The input of the braking prediction neural network model is the vehicle speed v0, the pedestrian's moving speed v1, the distance between the vehicle and the pedestrian d1, and the relative angle θ between the vehicle and the pedestrian. The output of the braking prediction neural network model is whether the vehicle brakes and the deceleration during braking.

[0032] A large amount of data is collected in advance from vehicles crossing zebra crossings without traffic lights. The data includes vehicle speed v0, pedestrian movement speed v1, distance d1 between the vehicle and the pedestrian, relative angle θ between the vehicle and the pedestrian, whether the vehicle brakes, and the deceleration of the vehicle when braking. The above data are used to train a braking prediction neural network model to obtain a trained braking prediction neural network model.

[0033] Step 2: While the vehicle is driving, the in-vehicle camera captures an image of the driver's face. The first image processor uses a machine learning model to determine whether the driver is distracted based on the facial image. The exterior camera captures an image of the road environment in front of the vehicle. The second image processor uses a machine vision method to identify the positions of pedestrians and zebra crossings based on the road environment image and determines whether a pedestrian is about to cross or is currently crossing the zebra crossing based on the pedestrian's movement direction and speed.

[0034] Step 3: When a pedestrian is about to cross the zebra crossing or is crossing the zebra crossing and the driver is distracted, and the distance between the vehicle and the pedestrian is less than or equal to 30m, record the vehicle speed v0 at the moment when the driver suddenly sees the pedestrian and is no longer distracted;

[0035] Step 4: The braking prediction neural network model predicts whether the vehicle will brake and the deceleration during braking based on the vehicle speed v0 at the moment when the driver suddenly sees the pedestrian and is no longer distracted, the pedestrian's moving speed v1, the distance d1 between the vehicle and the pedestrian, and the relative angle θ between the vehicle and the pedestrian.

[0036] Step 5: When it is predicted that the vehicle is about to brake, the processor controls the brake lights to light up;

[0037] When the distance between the vehicle and the following vehicle is less than 20m, and the vehicle is predicted to move at a speed of 3.5m / s 2 to 5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake lights will turn on immediately.

[0038] If the distance between the vehicle and the following vehicle is between 20m and 30m, and the vehicle is predicted to move at a speed of 2.5m / s 2 to 3.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.25s.

[0039] If the distance between the vehicle and the following vehicle is greater than 30m, and the vehicle is predicted to move at 0m / s 2 to 2.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.5s.

[0040] The average driver's reaction time in an emergency situation is between 0.3 and 1 second. This means that a distracted driver will take 0.3 to 1 second to brake after seeing a pedestrian crossing the road. Combined with the braking time, this means the vehicle's brake lights illuminate at least 0.5 seconds after the distracted driver sees the pedestrian. If the vehicle behind is close to the pedestrian at this time, a rear-end collision is likely to occur. The method of the present invention predicts that a vehicle is about to brake and preemptively illuminates the brake lights based on the distance between the vehicle and the following vehicle, providing an early warning to the following vehicle and reducing the risk of traffic accidents.

[0041] Although this specification has provided a detailed description of the present invention using general descriptions and specific embodiments, it will be apparent to those skilled in the art that modifications and improvements may be made based on the present invention. Therefore, such modifications and improvements, which do not depart from the spirit of the present invention, are intended to be within the scope of protection claimed herein.

Claims

1. A method for preventing rear-end collisions at a zebra crossing without a signal light, characterized in that: Includes a driver distraction perception module, an outside pedestrian perception module, a first millimeter-wave radar, a second millimeter-wave radar, and a processor; The driver distraction perception module includes an in-vehicle camera and a first image processor; The in-car camera is used to collect the driver's facial image; The first image processor is used to determine whether the driver is in a distracted state based on the driver's facial image; The external pedestrian sensing module includes an external camera and a second image processor; the external camera is used to capture images of the road environment in front of the vehicle; The second image processor is used to determine whether a pedestrian is about to cross the zebra crossing or is crossing the zebra crossing based on the road environment image; The first millimeter-wave radar is used to collect the distance between the vehicle and pedestrians about to cross the zebra crossing or currently crossing the zebra crossing, as well as the pedestrians' movement direction and speed; The second millimeter-wave radar is used to collect the distance between the vehicle and the following vehicle; The processor is used to predict whether the vehicle should brake and the degree of deceleration during braking, and is also used to control the vehicle's brake lights to light up; The following steps are involved: Step 1: Establish and train a braking prediction neural network model to obtain a trained braking prediction neural network model; the braking prediction neural network model inputs are vehicle speed, pedestrian movement speed, distance between the vehicle and the pedestrian, and relative angle between the vehicle and the pedestrian; the braking prediction neural network model outputs whether the vehicle brakes and the deceleration during braking; Step 2: While the vehicle is driving, the in-vehicle camera captures an image of the driver's face. The first image processor uses a machine learning model to determine whether the driver is distracted based on the facial image. The exterior camera captures an image of the road environment in front of the vehicle. The second image processor uses a machine vision method to identify the positions of pedestrians and zebra crossings based on the road environment image and determines whether a pedestrian is about to cross or is currently crossing the zebra crossing based on the pedestrian's movement direction and speed. Step 3: When a pedestrian is about to cross a zebra crossing or is crossing a zebra crossing and the driver is distracted, and the distance between the vehicle and the pedestrian is less than or equal to 30 meters, record the vehicle speed at the moment when the driver suddenly sees the pedestrian and is no longer distracted; Step 4: The braking prediction neural network model predicts whether the vehicle will brake and the deceleration rate when braking based on the vehicle speed, the pedestrian's moving speed, the distance between the vehicle and the pedestrian, and the relative angle between the vehicle and the pedestrian at the moment the driver suddenly sees the pedestrian and is no longer distracted. Step 5: When it is predicted that the vehicle is about to brake, the processor controls the brake lights to light up; When the distance between the vehicle and the following vehicle is less than 20m, and the vehicle is predicted to move at a speed of 3.5m / s 2 to 5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake lights will turn on immediately. If the distance between the vehicle and the following vehicle is between 20m and 30m, and the vehicle is predicted to move at a speed of 2.5m / s 2 to 3.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.25s. If the distance between the vehicle and the following vehicle is greater than 30m, and the vehicle is predicted to move at 0m / s 2 to 2.5m / s 2 If the vehicle decelerates at a deceleration rate between 0 and 1, the brake light will turn on after 0.5s.

2. The method for preventing rear-end collisions at a zebra crossing without a signal light according to claim 1, characterized in that: The training of the braking prediction neural network model is specifically to collect a variety of data in advance from a large number of vehicles passing through zebra crossings without traffic lights. The data includes vehicle speed, pedestrian movement speed, distance between vehicle and pedestrian, relative angle between vehicle and pedestrian, whether the vehicle brakes and the deceleration of the vehicle when braking; the braking prediction neural network model is trained with the above data to obtain a trained braking prediction neural network model.