Anti-collision method and equipment for road pedestrians and medium
Through deep learning technology, pedestrians are identified and collision probability is detected in combination with sensor data, reminder and buffering strategies are activated, which solves the safety problems of pedestrians and vehicles collisions, improves traffic safety and efficiency, and reduces the degree of injury.
Patent Information
- Application Number
- CN202510348515.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-24
- Publication Date
- 2025-05-13
AI Technical Summary
When pedestrians cross the road, especially during the right turn of the vehicle or during peak commuting periods, collisions between pedestrians and vehicles are prone to collisions, resulting in reduced traffic safety and may cause traffic congestion.
Deep learning technology is used to identify pedestrians from surveillance images collected by vehicle vision sensors, and combined with vehicle distance sensors and acceleration sensors, the distance and collision probability between vehicles and pedestrians are detected in real time. When a pedestrian is in a dangerous position or the probability of collision is high, start the driver reminder strategy, pedestrian reminder strategy and vehicle collision buffering strategy.
Through early warning and buffering strategies, the probability of accidents will be significantly reduced, the safety and efficiency of pedestrians will be improved, and the degree of injury of pedestrians will be reduced in the event of inevitable collisions.
Smart Images
Figure CN119975337A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of vehicle control technology, and in particular to a method, device and medium for preventing collision of road pedestrians. Background Art
[0002] With the increase in the number of vehicles in society, new challenges are brought to pedestrian road safety.
[0003] In current travel scenarios, collisions are mostly avoided by vehicle drivers observing the movements of pedestrians, or by pedestrians observing the movement of vehicles. However, there are still many cases of collisions when pedestrians cross the road and encounter vehicles turning right, or there are many cases of scrapes between vehicles and pedestrians during rush hour commuting in the morning and evening, which reduces the safety of pedestrians. In addition, when a collision occurs, the long process of handling the accident will also cause road traffic congestion, reducing the efficiency of vehicle and pedestrian traffic. Therefore, it is particularly important to provide early warning of collisions between vehicles and pedestrians. Summary of the invention
[0004] The present application provides a method, device and medium for preventing collision of pedestrians on a road, which are used to solve at least one of the above-mentioned technical problems.
[0005] This application adopts the following technical solutions:
[0006] In a first aspect, the present application provides a method for avoiding collision of road pedestrians, the method comprising: receiving monitoring images within a preset range in front of a vehicle, to the side of a vehicle and to the rear of a vehicle acquired by a vehicle vision sensor; identifying pedestrians from the monitoring images using deep learning technology; when the pedestrian is to the side of the vehicle and / or to the rear of the vehicle, using a vehicle distance sensor to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, initiating a driver reminder strategy and / or a pedestrian reminder strategy; when the pedestrian is in front of the vehicle, receiving a vehicle acceleration value detected by a vehicle acceleration sensor to predict the probability of collision between the vehicle and the pedestrian based on the vehicle acceleration value, and initiating a vehicle collision buffer strategy when the collision probability is greater than a preset probability threshold.
[0007] In one possible implementation of the present application, the driver reminder strategy includes at least one or more of seat vibration, steering wheel vibration, vehicle buzzer sounding, in-vehicle display screen pop-up prompts and in-vehicle voice prompts; the pedestrian reminder strategy includes at least starting the horn and / or the vehicle flashing lights; the vehicle collision buffering strategy includes at least starting the collision buffering device arranged on the engine hood and / or triggering the airbag arranged under the hood.
[0008] In one possible implementation of the present application, the vehicle vision sensor is installed at least in the vehicle cockpit and faces the front, side and rear of the vehicle, and includes at least one or more of an intelligent pedestrian monitoring camera, a radar and a lidar; the vehicle acceleration sensor is installed at least on the front of the vehicle bumper, and includes at least one or more of a piezoelectric acceleration sensor, a piezoresistive acceleration sensor and a point capacitance acceleration sensor.
[0009] In a possible implementation of the present application, the collision probability between the vehicle and the pedestrian is predicted based on the vehicle acceleration value, including: determining the collision situation between the vehicle and the pedestrian based on the vehicle driving direction and the pedestrian driving direction, the collision situation at least including a first collision situation in which the vehicle driving direction and the pedestrian driving direction are perpendicular, and a second collision situation in which the vehicle driving direction and the pedestrian driving direction are parallel; if the collision situation is the first collision situation, obtaining the average moving speed of the pedestrian, and calculating the collision probability through the vehicle acceleration value, the average moving speed of the pedestrian, and the distance between the vehicle and the pedestrian; if the collision situation is the second collision situation, obtaining the pedestrian swing area, and calculating the collision probability through the occupied area of the vehicle on the current driving path and the pedestrian swing area.
[0010] In a possible implementation of the present application, the collision probability is calculated through the vehicle acceleration value, the average moving speed of the pedestrian, and the distance between the vehicle and the pedestrian, including: calculating the time t1 required for the pedestrian to leave the vehicle-occupied area based on the average moving speed of the pedestrian and the current position of the pedestrian; predicting the time t2 required for the vehicle to reach the pedestrian's current position through the vehicle acceleration value, the vehicle's current speed and the distance between the vehicle and the pedestrian; and determining the collision probability through the difference between the required time t1 and the required time t2.
[0011] In a possible implementation of the present application, the collision probability is determined by the difference between the required time t1 and the required time t2, including: subtracting the required time t1 from the required time t2 to obtain the difference between the required time t1 and the required time t2; when the difference is within a preset difference range, determining that the collision probability is less than a preset probability threshold; when the difference is a negative value, or the difference is less than the minimum value of the preset difference range, determining that the collision probability is greater than a preset probability threshold.
[0012] In a possible implementation of the present application, the collision probability is calculated based on the occupied area of the vehicle on the current driving path and the pedestrian swing area, including: determining whether the pedestrian swing area exceeds the isolation line, and if so, determining the maximum occupied range of the pedestrian swing area; determining the minimum occupied range of the vehicle on the current driving path of the vehicle; determining whether there is an overlapping area between the maximum occupied range and the minimum occupied range, and if not, determining that the collision probability is less than a preset probability threshold.
[0013] In a possible implementation of the present application, the method also includes: determining an overlapping area between the maximum occupancy range and the minimum occupancy range; predicting the time t3 when the vehicle arrives at the overlapping area based on the vehicle acceleration value, and predicting the time t4 when the pedestrian arrives at the overlapping area based on the average moving speed of the pedestrian; if the difference between the time t3 and the time t4 is less than a preset difference threshold, determining that the collision probability is less than a preset probability threshold.
[0014] In a second aspect, the present application also provides a collision avoidance device for road pedestrians, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor so that the at least one processor can execute: receiving monitoring images within a preset range in front of the vehicle, on the side of the vehicle and behind the vehicle acquired by a vehicle visual sensor; identifying pedestrians from the monitoring images using deep learning technology; when the pedestrian is on the side of the vehicle and / or behind the vehicle, using a vehicle distance sensor to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, starting a driver reminder strategy and / or a pedestrian reminder strategy; when the pedestrian is in front of the vehicle, receiving a vehicle acceleration value detected by a vehicle acceleration sensor to predict the probability of collision between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, starting a vehicle collision buffer strategy.
[0015] In the third aspect, the present application also provides a non-volatile computer storage medium having computer executable instructions stored thereon, wherein the computer executable instructions are configured to execute: receiving monitoring images within a preset range in front of the vehicle, on the side of the vehicle and behind the vehicle acquired by a vehicle vision sensor; identifying pedestrians from the monitoring images using deep learning technology; when the pedestrian is on the side of the vehicle and / or behind the vehicle, using a vehicle distance sensor to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, initiating a driver reminder strategy and / or a pedestrian reminder strategy; when the pedestrian is in front of the vehicle, receiving a vehicle acceleration value detected by a vehicle acceleration sensor to predict the probability of collision between the vehicle and the pedestrian based on the vehicle acceleration value, and initiating a vehicle collision buffer strategy when the collision probability is greater than a preset probability threshold.
[0016] The present application provides a method, device and medium for preventing collision of pedestrians on a road, which have the following beneficial effects:
[0017] 1) Reduce accident risks: By monitoring the position and speed of pedestrians and vehicles in real time, early warnings can be issued before potential collisions occur, reminding drivers and pedestrians to take evasive measures, thereby significantly reducing the probability of accidents. In addition, the reminder strategy can also urge pedestrians to pass as quickly as possible, improving the safety and efficiency of pedestrian traffic.
[0018] 2) Reduce the degree of injury: In the event of an unavoidable collision, the vehicle's collision buffering strategies can be pre-activated, such as a pop-up hood, to reduce the risk of pedestrians' head and leg injuries, that is, to reduce the degree of injury to pedestrians, thereby reducing the losses caused by the collision. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the technical solutions in this application or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:
[0020] Figure 1 A flow chart of a method for avoiding collision of pedestrians on a road provided in this application;
[0021] Figure 2 A schematic structural diagram of a road pedestrian anti-collision device provided in this application. DETAILED DESCRIPTION
[0022] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in this application will be described clearly and completely below in conjunction with the drawings in this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.
[0023] The method in this application is described in detail below with reference to the accompanying drawings.
[0024] Figure 1 A flowchart of a method for avoiding collision of pedestrians on a road provided in this application, such as Figure 1 As shown, the road pedestrian collision avoidance method in this application at least includes the following execution steps:
[0025] Step 101: Receive monitoring images within a preset range in front of, to the side of, and behind a vehicle collected by a vehicle visual sensor.
[0026] The pedestrian collision avoidance method in the present application is applied to a vehicle, and its execution subject may be a server of the vehicle's whole vehicle control system.
[0027] First, the vehicle visual sensor is used to collect monitoring images within a preset range in front of the vehicle, on the side of the vehicle, and behind the vehicle. For example, monitoring images within a range of 100 meters in front of the vehicle and 10 meters behind the vehicle are collected.
[0028] Of course, video stream data can also be collected here to obtain monitoring images through the video stream data.
[0029] In one possible implementation of the present application, the vehicle vision sensor in the present application includes at least one of a camera, a radar and a lidar, and can be installed in the vehicle cockpit facing the front, side and rear of the vehicle to detect pedestrians and track their movements.
[0030] For example, the visual sensor in this application can use the Uniview intelligent capture unit: Uniview Technology's intelligent capture unit uses advanced image fusion technology, combined with deep intelligent algorithms, to output high-quality full-color images in low-light environments (such as under city street lights). Products include environmental protection checkpoint capture units, environmental protection electric police capture units, toll station checkpoint capture units, etc. You can also use the STONKAM 1080P intelligent high-definition real-time pedestrian monitoring and warning camera: STONKAM's 1080P intelligent high-definition real-time pedestrian monitoring and warning camera uses deep learning technology to intelligently detect pedestrians in front, on the side and behind the vehicle in real time. When used with STONKAM high-definition displays, it can also provide real-time audio and video alarms to promptly remind drivers of potential risks and avoid collisions with pedestrians.
[0031] Step 102: Use deep learning technology to identify pedestrians from the surveillance image.
[0032] After obtaining the monitoring image, the pedestrians therein can be identified by the visual sensor itself, or the monitoring image can be transmitted to the server of the vehicle control system or the monitoring server for pedestrian identification. In addition, the process of identifying or detecting pedestrians in the monitoring image can be implemented by existing deep learning technology or deep learning algorithm, which is not described in detail in the embodiments of the present application.
[0033] Step 103: When the pedestrian is at the side of the vehicle and / or behind the vehicle, the vehicle distance sensor is used to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, the driver reminder strategy and / or the pedestrian reminder strategy is activated.
[0034] If the visual sensor determines that a pedestrian is on the side or behind the vehicle, the distance sensor and other technologies will be used to measure the distance between the vehicle and the pedestrian. If the distance between the two is less than the preset distance threshold, it means that there is a risk of pedestrian collision. At this time, the driver reminder strategy and / or pedestrian reminder strategy will be activated.
[0035] In a possible implementation of the present application, the driver reminder strategy includes at least one or more of vibrating the seat, vibrating the steering wheel, sounding the vehicle buzzer, pop-up window prompts on the in-vehicle display screen, and in-vehicle voice prompts; the pedestrian reminder strategy includes at least starting the horn and / or the vehicle flashing lights. By reminding the driver through the driver reminder strategy, or reminding the pedestrian through the pedestrian reminder strategy, the occurrence of traffic accidents can be effectively reduced, and pedestrians can be urged to pass, thereby improving the efficiency and safety of passage.
[0036] Step 104: When a pedestrian is in front of the vehicle, a vehicle acceleration value detected by a vehicle acceleration sensor is received to predict the collision probability between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, a vehicle collision buffer strategy is initiated.
[0037] If the pedestrian is detected in front of the vehicle by the visual sensor, the vehicle acceleration is detected by the vehicle acceleration sensor installed at the front end of the vehicle bumper. Specifically, the vehicle acceleration sensor here includes at least one or more of a piezoelectric acceleration sensor, a piezoresistive acceleration sensor and a point capacitance acceleration sensor. In one example, the principle of the piezoelectric acceleration sensor is to use the piezoelectric effect of piezoelectric materials (such as piezoelectric ceramics or quartz crystals). When the piezoelectric material is subjected to pressure or vibration, charges will be generated inside it, and the value of acceleration can be obtained by measuring these charges. It is characterized by high sensitivity, high frequency response and good linearity, suitable for high-precision measurement, and is widely used in automobile collision experiments, equipment vibration monitoring and other fields. The principle of the piezoresistive acceleration sensor is based on the piezoresistive effect of semiconductor materials. When the semiconductor material is subjected to pressure, its resistance value will change, and the value of acceleration can be obtained by measuring the change in resistance value. It is characterized by small size, low power consumption, and easy integration in various analog and digital circuits. It is also widely used in automobile collision experiments, test instruments, equipment vibration monitoring and other fields. The principle of capacitive acceleration sensor is to use the change of capacitance to measure acceleration. When the sensor is subjected to acceleration, the internal capacitance value will change, and the acceleration value can be obtained by measuring these changes. Its characteristics are high sensitivity, low noise, good linearity and wide measurement range. It is also used in measurement scenarios that require high precision.
[0038] In one example, a touch sensor may also be installed at the front end of the vehicle bumper. The vehicle acceleration sensor or touch sensor detects that the distance between the vehicle and the pedestrian is too small. When a collision occurs, a signal is immediately sent to the control system, which allows the control system to take measures such as emergency braking to reduce the intensity of the collision. When the vehicle cannot avoid a collision, the vehicle will activate the buffer device of the engine hood at the moment the collision is detected to reduce the collision damage to the pedestrian. Some vehicles are also equipped with airbags, which will be triggered when the hood is opened, thereby reducing pedestrian injuries by increasing energy absorption and offsetting the impact force along the surface of the engine hood. This solution is deployed in the vehicle control system as a pedestrian protection strategy.
[0039] In a possible implementation of the present application, after obtaining the vehicle acceleration value, the collision probability between the vehicle and the pedestrian can also be predicted by the vehicle acceleration value. Specifically, first, the driving direction of the vehicle and the driving direction of the pedestrian are determined. The driving direction of the vehicle can be obtained by the vehicle control system, and the driving direction of the pedestrian can be determined by the aforementioned visual sensor. If the driving direction of the vehicle and the driving direction of the pedestrian are in vertical directions, then when the two collide, it is a first collision situation. If the driving direction of the vehicle and the driving direction of the pedestrian are parallel, then when the two collide, it is a second collision situation.
[0040] For the first collision scenario, the average moving speed, current position of the pedestrian, and the distance between the vehicle and the pedestrian are obtained. This information can be obtained through monitoring images or monitoring video streams collected by visual sensors based on the distance sensor installed on the vehicle. Afterwards, based on the average moving speed and current position of the pedestrian, the time t1 required for the pedestrian to leave the vehicle occupied area is predicted. The vehicle occupied area is the occupied area in the direction of the pedestrian's travel after the vehicle continues to move forward and reaches the pedestrian's current position. In one example, the occupied area of the vehicle is the width of the vehicle. The time t2 required for the vehicle to reach the pedestrian's current position is calculated based on the vehicle's current speed, acceleration value, and distance between the vehicle and the pedestrian. If the time t2 required by the vehicle is greater than the time t1 required by the pedestrian, it means that the pedestrian leaves the vehicle occupied area first and the vehicle arrives at the pedestrian's current position later. At this time, the two will not collide. If the time t2 required by the vehicle is less than the time t1 required by the pedestrian, it means that the vehicle arrives at the pedestrian's current position first and the pedestrian leaves the vehicle occupied area later. At this time, a collision is very likely to occur. Therefore, in the embodiment of the present application, the difference between the time t2 required for the vehicle and the time t1 required for the pedestrian is determined. When the difference is within the preset difference range, it is determined that the probability of a collision is less than the preset probability threshold, and the pedestrian can pass safely. If the difference is a negative value, or the difference is less than the minimum value of the preset difference range, it means that the vehicle arrives first, or the time required for the vehicle and the pedestrian are close. At this time, since the pedestrian is not moving at a uniform speed, a collision is very likely to occur. Therefore, in this case, the probability of a collision is determined to be greater than the preset probability threshold.
[0041] For the second collision situation, since pedestrians do not travel in a straight line when driving forward, they may swing or sway left and right due to the influence of other vehicles or other roadblocks. At this time, the range of the pedestrian's swinging left and right is determined as the pedestrian's swinging area. First, determine whether the pedestrian's swinging area exceeds the isolation line, that is, the guide line on the road surface that separates vehicles and pedestrians. If it does not exceed, it means that the two will not collide. If it exceeds, determine the maximum occupied range of the pedestrian's swinging area. The maximum occupied range here refers to the maximum range of the part that exceeds the isolation line. Then, on the current driving path of the vehicle, determine the minimum occupied range of the vehicle, because in general, when the pedestrian deviates or swings toward the lane of the vehicle, the vehicle can choose to avoid it. The occupied range after the vehicle avoids is the minimum occupied range. If there is no overlapping area between the maximum occupied range of the pedestrian and the minimum occupied range of the vehicle, it is determined that the probability of a collision between the two is greater than the preset probability threshold, that is, the two will not collide. However, if there is an overlapping area between the two, the time t3 when the vehicle arrives at the overlapping area is predicted based on the vehicle acceleration value, and the time t4 when the pedestrian arrives at the overlapping area is predicted based on the average moving speed of the pedestrian, and the time difference between t3 and t4 is calculated. If the time difference is large, it means that the interval between the two arriving at the overlapping area is large, and no collision will occur at this time. However, if the time difference between the two is less than the preset difference threshold, that is, the two will arrive at the overlapping area at a similar time, then it is determined that the probability of a collision is less than the preset probability threshold, that is, a collision will occur.
[0042] Furthermore, whether for the first collision scenario or the second collision scenario, when it is determined that a collision may occur between the vehicle and the pedestrian, in order to ensure the safety of the pedestrian and minimize the collision loss, the aforementioned vehicle collision buffer strategy is activated, that is, the collision buffer device arranged on the engine hood is activated and / or the airbag arranged under the hood is triggered.
[0043] That is, through the above method, timely detection and warning of collisions between vehicles and pedestrians can be carried out, which can effectively reduce the occurrence of collisions and improve the traffic efficiency and safety of vehicles and pedestrians. On the other hand, when a collision occurs, the vehicle collision buffer strategy can be activated to effectively reduce the degree of injury to pedestrians and reduce the losses caused by the collision.
[0044] Based on the same inventive concept, the present application also provides a road pedestrian anti-collision device, the structure of which is as follows: Figure 2 shown.
[0045] Figure 2 This is a schematic diagram of the structure of a road pedestrian anti-collision device provided in this application. Figure 2As shown, the road pedestrian collision avoidance device 200 in the present application specifically includes: at least one processor 201; and, a memory 203 that is communicatively connected to the at least one processor 201 (connected via a bus 202); wherein the memory 203 stores instructions that can be executed by the at least one processor 201, so that the at least one processor 201 can execute a road pedestrian collision avoidance method as described in any of the above embodiments.
[0046] In one possible implementation of the present application, the aforementioned processor is used to execute, receiving monitoring images within a preset range in front of the vehicle, on the side of the vehicle and behind the vehicle collected by the vehicle vision sensor; using deep learning technology to identify pedestrians from the monitoring images; when the pedestrian is on the side of the vehicle and / or behind the vehicle, using the vehicle distance sensor to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, starting the driver reminder strategy and / or the pedestrian reminder strategy; when the pedestrian is in front of the vehicle, receiving the vehicle acceleration value detected by the vehicle acceleration sensor to predict the collision probability between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, starting the vehicle collision buffer strategy.
[0047] In addition, the present application also provides a non-volatile computer storage medium on which computer executable instructions are stored, and the computer executable instructions are configured to execute a road pedestrian collision avoidance method as described in any of the above embodiments.
[0048] In one possible implementation of the present application, the aforementioned computer executable instructions are configured to execute: receiving monitoring images within a preset range in front of the vehicle, on the side of the vehicle and behind the vehicle acquired by a vehicle vision sensor; identifying pedestrians from the monitoring images using deep learning technology; when the pedestrian is on the side of the vehicle and / or behind the vehicle, using a vehicle distance sensor to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, initiating a driver reminder strategy and / or a pedestrian reminder strategy; when the pedestrian is in front of the vehicle, receiving a vehicle acceleration value detected by a vehicle acceleration sensor to predict the probability of collision between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, initiating a vehicle collision buffer strategy.
[0049] Of course, the solution in this application can also be proposed as a system, which uses radars, cameras, etc. to fully perceive the running conditions of vehicles and pedestrians on the road, and acts as an information aggregation base station. The control system issues instructions to the traffic guidance screen for early warning and notification, and uses horns and flashing lights to warn passing vehicles to drive carefully. A set of equipment is deployed in each direction of the intersection, and multiple devices form a collaborative system to remind oncoming vehicles to control speed and the danger ahead, and at the same time notify the other two parties to pay attention to safety.
[0050] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0051] The device and method provided in the present application correspond one to one, and therefore, the device also has similar beneficial technical effects as its corresponding method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the device will not be repeated here.
[0052] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media that include computer-usable program code.
[0053] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0054] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.
Claims
1. A method for preventing collision of pedestrians on a road, characterized in that: The method comprises: Receiving monitoring images within a preset range in front of, to the side of, and behind the vehicle collected by the vehicle visual sensor; Using deep learning technology to identify pedestrians from the surveillance image; When the pedestrian is at the side of the vehicle and / or behind the vehicle, the vehicle distance sensor is used to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, the driver reminder strategy and / or the pedestrian reminder strategy is activated; When the pedestrian is in front of the vehicle, a vehicle acceleration value detected by a vehicle acceleration sensor is received to predict the collision probability between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, a vehicle collision buffer strategy is initiated.
2. A method for preventing collision of pedestrians on a road according to claim 1, characterized in that: The driver reminder strategy includes at least one or more of seat vibration, steering wheel vibration, vehicle buzzer sound, in-vehicle display screen pop-up window prompt and in-vehicle voice prompt; the pedestrian reminder strategy includes at least activating the horn and / or vehicle flashing lights; The vehicle collision buffering strategy at least includes activating a collision buffering device disposed on the engine hood and / or triggering an airbag disposed under the engine hood.
3. A method for preventing collision of pedestrians on a road according to claim 1, characterized in that: The vehicle visual sensor is installed at least in the vehicle cockpit and faces the front, side and rear of the vehicle, and includes at least one or more of an intelligent pedestrian monitoring camera, a radar and a laser radar; The vehicle acceleration sensor is at least installed on the front part of the vehicle bumper, and includes at least one or more of a piezoelectric acceleration sensor, a piezoresistive acceleration sensor and a point capacitance acceleration sensor.
4. A method for preventing collision of pedestrians on a road according to claim 1, characterized in that: Predicting the collision probability between the vehicle and the pedestrian according to the vehicle acceleration value includes: Determining a collision situation between the vehicle and the pedestrian according to the driving direction of the vehicle and the driving direction of the pedestrian, wherein the collision situation includes at least a first collision situation in which the driving direction of the vehicle and the driving direction of the pedestrian are perpendicular, and a second collision situation in which the driving direction of the vehicle and the driving direction of the pedestrian are parallel; If the collision situation is the first collision situation, the average moving speed of the pedestrian is obtained, and the collision probability is calculated according to the vehicle acceleration value, the average moving speed of the pedestrian, and the distance between the vehicle and the pedestrian; If the collision situation is the second collision situation, the pedestrian swing area is obtained, and the collision probability is calculated through the occupied area of the vehicle on the current driving path and the pedestrian swing area.
5. A method for preventing collision of pedestrians on a road according to claim 4, characterized in that: The collision probability is calculated by using the vehicle acceleration value, the average moving speed of the pedestrian, and the distance between the vehicle and the pedestrian, including: Calculate the time t1 required for the pedestrian to leave the vehicle occupied area based on the average moving speed of the pedestrian and the current location of the pedestrian; Predicting the time t2 required for the vehicle to reach the current position of the pedestrian based on the vehicle acceleration value, the current speed of the vehicle and the distance between the vehicle and the pedestrian; The collision probability is determined by the difference between the required time t1 and the required time t2.
6. A method for preventing collision of pedestrians on a road according to claim 5, characterized in that: Determining the collision probability by the difference between the required time t1 and the required time t2 includes: Subtracting the required time t1 from the required time t2 to obtain a difference between the required time t1 and the required time t2; When the difference is within a preset difference range, determining that the collision probability is less than a preset probability threshold; When the difference is a negative value, or the difference is smaller than the minimum value of the preset difference range, it is determined that the collision probability is greater than a preset probability threshold.
7. A method for preventing pedestrians from collision according to claim 4, characterized in that: The collision probability is calculated based on the occupied area of the vehicle on the current driving path and the pedestrian swing area, including: Determining whether the pedestrian swing area exceeds the isolation line, and if so, determining the maximum occupied range of the pedestrian swing area; Determine the minimum occupancy range of the vehicle on the current driving path of the vehicle; It is determined whether there is an overlapping area between the maximum occupied range and the minimum occupied range, and if not, it is determined that the collision probability is less than a preset probability threshold.
8. A method for preventing collision of pedestrians on a road according to claim 7, characterized in that: The method further comprises: Determine an overlapping area between the maximum occupied range and the minimum occupied range; Predicting a time t3 when the vehicle arrives at the overlapping area based on the vehicle acceleration value, and predicting a time t4 when the pedestrian arrives at the overlapping area based on the average moving speed of the pedestrian; If the difference between the time t3 and the time t4 is less than a preset difference threshold, it is determined that the collision probability is less than a preset probability threshold.
9. A collision avoidance device for pedestrians on a road, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to: Receiving monitoring images within a preset range in front of, to the side of, and behind the vehicle collected by the vehicle visual sensor; Using deep learning technology to identify pedestrians from the surveillance image; When the pedestrian is at the side of the vehicle and / or behind the vehicle, the vehicle distance sensor is used to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, the driver reminder strategy and / or the pedestrian reminder strategy is activated; When the pedestrian is in front of the vehicle, a vehicle acceleration value detected by a vehicle acceleration sensor is received to predict the collision probability between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, a vehicle collision buffer strategy is initiated.
10. A non-volatile computer storage medium having computer executable instructions stored thereon, characterized in that: The computer executable instructions are configured to perform: Receiving monitoring images within a preset range in front of, to the side of, and behind the vehicle collected by the vehicle visual sensor; Using deep learning technology to identify pedestrians from the surveillance image; When the pedestrian is at the side of the vehicle and / or behind the vehicle, the vehicle distance sensor is used to detect the distance between the vehicle and the pedestrian in real time, and when the distance is less than a preset distance threshold, the driver reminder strategy and / or the pedestrian reminder strategy is activated; When the pedestrian is in front of the vehicle, a vehicle acceleration value detected by a vehicle acceleration sensor is received to predict the collision probability between the vehicle and the pedestrian based on the vehicle acceleration value, and when the collision probability is greater than a preset probability threshold, a vehicle collision buffer strategy is initiated.