A green driver communication system based on machine vision and vehicle network

Abstract

The present invention discloses a green driver communication system based on machine vision and vehicle network, including a communication device, the communication device including a depth camera for collecting depth images, a DSP module for processing depth images, a vehicle network module for signal broadcasting and signal reception, a sound alarm module for being installed in the vehicle and alerting the driver that a vehicle is entering from behind, and an LCD display module for displaying the position of the vehicle from which the signal comes, the depth camera is connected to the DSP module, and the DSP module is connected to the vehicle network module, the sound alarm module and the LCD display module. The green driver communication system of the present invention can effectively realize mutual communication between drivers, is more accurate in the objects it faces, can display the position of the vehicle from which the signal comes, can alert the driver that a vehicle is entering from behind through the sound in the vehicle, is more efficient, and is a green driver communication system that is more environmentally friendly.

Description

Technical Field

[0001] The present invention relates to the fields of computer vision and vehicle mobile communication, and in particular to a green driver communication system based on machine vision and a vehicle-mounted network. Background Art

[0002] With the rapid growth in the number of cars and the fact that the phenomenon of random honking has not been curbed because it is difficult to define, the noise pollution caused by car honking is becoming more and more serious.

[0003] In response to the above problems, the invention patent with application number CN201720029031.9 discloses an interactive silent whistle system for vehicles, whose main components include a single chip microcomputer, a transmitting component, a receiving component, a Bluetooth module component, an audio component, a memory, etc., and the single chip microcomputer is connected to other modules respectively. This interactive silent whistle system can reduce the noise pollution of car whistles to a certain extent, but its practicality needs to be improved. The main disadvantages are: first, the above-mentioned interactive silent whistle system for vehicles uses Bluetooth broadcasting to propagate the whistle signal. Vehicles within a certain range receive the signal and trigger the in-car audio, and the signal orientation is still not accurate enough; second, the triggering of its device still depends on the manual operation of the driver, and the automatic triggering of the whistle device cannot be realized; third, when there are multiple signal sources, it fails to realize the priority sorting of signal responses; fourth, its signal display form is only one kind of audio device, which is relatively single.

[0004] In summary, this interactive silent whistle system for vehicles is not applicable to current driving conditions, and a new system needs to be invented for green communication between drivers. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a green driver communication system based on machine vision and vehicle-mounted network in response to the deficiencies of the above-mentioned prior art. The green driver communication system based on machine vision and vehicle-mounted network can effectively realize mutual communication between drivers, is more accurate in the objects it faces, can display the location of the vehicle from which the signal comes, and can warn the driver of a vehicle entering from behind through sound in the car; it is more efficient; and it is a green driver communication system that is more environmentally friendly.

[0006] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0007] A green driver communication system based on machine vision and vehicle network includes a communication device, wherein the communication device includes a depth camera for collecting depth images, a DSP module for processing the depth images, an on-board network module for signal broadcasting and signal reception, a sound alarm module for being installed in the vehicle and alerting the driver that a vehicle is approaching from behind, and an LCD display module for displaying the position of the vehicle from which the signal is sourced, wherein the depth camera is connected to the DSP module, and the DSP module is connected to the on-board network module, the sound alarm module and the LCD display module.

[0008] As a further improved technical solution of the present invention, it also includes a power module, an electronic switch and a main controller, the power module is connected to the AC device through the electronic switch, the main controller is connected to the vehicle speed sensor on the vehicle, the main controller is connected to the electronic switch, and the main controller is used to control the on and off of the electronic switch according to the vehicle speed detected by the speed sensor, thereby controlling the power on and off of the AC device.

[0009] As a further improved technical solution of the present invention, the depth camera is arranged at the front windshield of the vehicle, at the left and right rearview mirrors, or in front of the rearview mirror inside the vehicle.

[0010] As a further improved technical solution of the present invention, the depth camera is used to periodically collect depth image information of the target vehicle.

[0011] As a further improved technical solution of the present invention, in each working cycle, when the depth camera collects depth image information of the target vehicle, the DSP module is used to obtain the license plate number string of the target vehicle from the depth image information based on the license plate number recognition system; the DSP module is used to obtain the position information of the target vehicle relative to the present vehicle from the depth image information based on the principle of the depth camera, and the position information includes the lateral distance of the target vehicle relative to the present vehicle, the longitudinal distance of the target vehicle relative to the present vehicle, and the direction of the target vehicle relative to the present vehicle; the DSP module is used to calculate the longitudinal speed of the target vehicle relative to the present vehicle based on the position information obtained from two frames of depth images; the DSP module is used to calculate the road width based on the collected depth image information.

[0012] As a further improved technical solution of the present invention, the DSP module is used to screen out the real target vehicle according to the road width and the position information of each target vehicle in the target vehicle group relative to the present vehicle, and calculate the relative collision time between the real target vehicle and the present vehicle according to the position information and longitudinal speed of the real target vehicle relative to the present vehicle, and then divide the collision risk level according to the relative collision time.

[0013] As a further improved technical solution of the present invention, the DSP module is used to determine whether there is a collision risk based on the collision risk level. If so, the DSP module integrates and encodes the license plate number string information of the real target vehicle, the license plate number string information of the vehicle, the position information of the vehicle relative to the real target vehicle, and the collision risk level information in sequence based on the Unicode encoding method.

[0014] As a further improved technical solution of the present invention, the vehicle network module is used to broadcast the encoded information using the DSRC protocol, and all vehicles within the broadcast range can receive the information through the vehicle network module installed on their own vehicles. The DSP module installed on the vehicle that receives the information is used to decode the received information based on the Unicode encoding method. The DSP module is used to compare and verify the license plate number string of its own vehicle with the first item in the decoded information. When the result of the comparison and verification is the same, the decoded information passes the verification; if there is only one piece of information that passes the verification, the buzzer in the sound alarm module installed on the vehicle that passes the verification is used to generate a signal according to the collision risk, etc. in the information that passes the verification. The DSP module installed on the verified vehicle is used to select the highest collision risk level from all the verified information, and the buzzer in the sound alarm module installed on the verified vehicle is used to issue a sound warning at a frequency corresponding to the highest collision risk level. At the same time, the LCD display module installed on the verified vehicle is used to display the positions of all signal source vehicles according to the color interface corresponding to the collision risk level in the decoded information.

[0015] The beneficial effects of the present invention are:

[0016] (1) The present invention utilizes a depth camera for image acquisition. Not only can the position information of the target vehicle relative to the signal source vehicle be effectively extracted from the acquired image information, but the DSP module can also recognize the license plate number through the acquired depth image, thereby encoding the signal according to the recognized license plate number, achieving multiple goals at one stroke.

[0017] (2) Compared with radar and ultrasonic sensors, the depth camera used in the present invention has the advantages of wider viewing angle, larger monitoring range and longer detection distance, thus providing drivers with longer reaction time, thereby reducing the accident rate;

[0018] (3) Compared with the traditional whistle device, the signal propagation speed of the present invention is faster. The traditional whistle device transmits signals through sound, and its signal propagation speed is about v sound=340m / s, while the speed of signal propagation through wireless networks (essentially electromagnetic waves) is close to the speed of light c. On a highway, if the distance between vehicles is the minimum limit distance s=100m, the signal transmission method provided by the present invention can save about the following time compared with the traditional whistle: This provides the driver with longer reaction time, making high-speed driving safer.

[0019] (4) The present invention utilizes a vehicle-mounted self-organizing network for signal transmission, which reduces noise pollution and is more environmentally friendly than conventional whistle devices that transmit signals through sound.

[0020] (5) Compared with the traditional horn device, the present invention can still work effectively in no-horn zones such as suburban roads without violating policy requirements, thereby improving the communication efficiency between drivers.

[0021] (6) The present invention uses computer vision technology to collect images of vehicles (target vehicles) that hinder the normal driving of the vehicle (i.e., the signal source vehicle), and based on this, obtains the license plate number of the target vehicle as an address identifier, thereby achieving accurate signal delivery and a sufficiently precise signal orientation; the harmful impact range of the signal is smaller, and it is more environmentally friendly.

[0022] (7) The early warning reminder strategy of the present invention adopts a combination of audio and video, which can quickly mobilize the driver's vigilance when the risk of collision is high, thereby improving safety.

[0023] (8) The present invention reasonably divides the danger levels of possible collisions based on the estimation of the collision time TTC, so that the buzzer in the sound alarm module installed on the verified vehicle emits a sound warning at a frequency corresponding to the highest collision danger level in all decoded information. At the same time, the LCD display screen in the LCD display module installed on the verified vehicle displays the positions of all signal source vehicles in an interface with a color corresponding to the collision danger level in the decoded information, and different colors are used to represent different collision danger levels. The positions of all signal source vehicles are displayed in color, so that the driver can intuitively and quickly prioritize emergencies, thereby improving driving safety; the human-computer interaction is good.

[0024] (9) The master controller of the present invention is connected to the vehicle speed sensor on the vehicle and is used to control the on and off of the electronic switch according to the vehicle speed detected by the vehicle speed sensor. When the vehicle speed sensor detects that the vehicle speed is 0, the master controller controls the electronic switch to be turned off and the AC device does not work; when the vehicle speed sensor detects that the vehicle speed is not 0, the master controller controls the electronic switch to be turned on and the AC device is powered on to work. Therefore, when the vehicle is driving, the AC device is automatically triggered and starts working without the need for manual operation by the driver.

[0025] (10) The present invention is a "non-complete broadcast" green horn system that can effectively realize mutual communication between drivers, realize automatic identification of horn signal targets, target more accurately, display the location of the signal source vehicle and the collision risk level by color, has higher efficiency, good human-computer interaction performance, and is more environmentally friendly. The system is applicable to highways and suburban roads. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a structural schematic diagram of the present invention.

[0027] Figure 2 Schematic diagram of several situations where the whistle needs to be blown that are considered by the present invention:

[0028] in Figure 2 (a) is a schematic diagram of the situation when there is a car in front of the left when overtaking;

[0029] in Figure 2 (b) is a schematic diagram of the situation when there is a car in front of the right when overtaking;

[0030] in Figure 2 Middle (c) is a schematic diagram of the situation when there are cars both in front of the left and right when overtaking;

[0031] in Figure 2 The middle (d) is a schematic diagram of the situation when there are cars in front of the left, right and front when overtaking.

[0032] Figure 3 The figure is a working flow chart of the structure of the present invention in one cycle.

[0033] Figure 4 Screening of real target vehicle schematics for the present invention.

[0034] Figure 5 Schematic diagram of the position of the signal source vehicle relative to the real target vehicle of the present invention. DETAILED DESCRIPTION

[0035] The specific embodiments of the present invention are further described below according to the accompanying drawings:

[0036] like Figure 1As shown, a green driver communication system based on machine vision and vehicle network includes a communication device, the communication device includes a depth camera for collecting depth images, a DSP module for processing depth images, a vehicle network module for signal broadcasting and signal reception, a sound alarm module for being installed in the vehicle and alerting the driver that a vehicle is approaching from behind, and an LCD display module for displaying the license plate number and position of the vehicle from which the signal comes, the depth camera is connected to the DSP module, and the DSP module is connected to the vehicle network module, the sound alarm module and the LCD display module. The LCD display module uses an LCD display screen, and the sound alarm module uses a buzzer. A green driver communication system is installed on each vehicle.

[0037] The above-mentioned green driver AC system also includes a power module, an electronic switch and a main controller. The power module is connected to the AC device through the electronic switch, the main controller is connected to the vehicle speed sensor on the vehicle, and the main controller is connected to the electronic switch. The main controller is used to control the on and off of the electronic switch according to the vehicle speed detected by the speed sensor, thereby controlling the power on and off of the AC device.

[0038] The depth camera of this embodiment is arranged at the front windshield of the vehicle, at the left and right rearview mirrors, or in front of the rearview mirror inside the vehicle.

[0039] like Figure 1 As shown, the depth camera of this embodiment periodically collects depth image information of the target vehicle. In each working cycle, when the depth camera collects the depth image information of the target vehicle, the DSP module obtains the license plate number string of the target vehicle from the depth image information based on the license plate number recognition system; the DSP module obtains the position information of the target vehicle relative to the vehicle from the depth image information based on the principle of the depth camera, and the position information of the target vehicle relative to the vehicle includes the lateral distance of the target vehicle relative to the vehicle, the longitudinal distance of the target vehicle relative to the vehicle, and the direction of the target vehicle relative to the vehicle; the DSP module calculates the longitudinal speed of the target vehicle relative to the vehicle based on the position information obtained from the two frames of depth images; the DSP module calculates the road width based on the collected depth image information.

[0040] The DSP module of this embodiment also screens out the real target vehicles based on the road width and the lateral distance of each target vehicle in the target vehicle group relative to the present vehicle, and calculates the relative collision time between the real target vehicle and the present vehicle based on the longitudinal distance and longitudinal speed of the real target vehicle relative to the present vehicle, and then divides the collision risk level according to the relative collision time.

[0041] The DSP module also determines whether there is a collision risk based on the collision risk level. That is, when the collision risk level is level one, level two or level three, there is a collision risk. If there is a collision risk, the DSP module integrates and encodes the license plate number string information of the real target vehicle, the license plate number string information of this vehicle, the position information of this vehicle relative to the real target vehicle, and the collision risk level information in sequence based on the Unicode encoding method.

[0042] The vehicle network module of this embodiment broadcasts the encoded information using the DSRC protocol. All vehicles within the broadcast range can receive the information through the vehicle network module installed on their own vehicles. The DSP module installed on the vehicle that receives the information decodes the received information based on the Unicode encoding method. The DSP module compares and verifies the license plate number string of its own vehicle with the first item in the decoded information. When the result of the comparison and verification is the same, the decoded information passes the verification, that is, passes the identity authentication. If there is one piece of information that passes the verification, the sound alarm module installed on the verified vehicle is under the control of the DSP module. The buzzer in the sound alarm module emits a sound warning at a frequency corresponding to the collision risk level in the verified information. The LCD display module installed on the verified vehicle displays the license plate number and position of the signal source vehicle in an interface with a color corresponding to the collision risk level in the decoded information. If there are multiple pieces of information that have passed the verification, that is, when signals broadcast by multiple vehicles are received at the same time, the DSP module selects the one with the highest collision risk level from all the information that has passed the verification, and controls the sound alarm module according to the highest collision risk level. The buzzer in the sound alarm module installed on the verified vehicle emits a sound warning at a frequency corresponding to the highest collision risk level. At the same time, the LCD display module installed on the verified vehicle displays the license plate number and position of all signal source vehicles in an interface with a color corresponding to the collision risk level in the verified information under the control of the DSP module. Different sound frequencies correspond to different collision risk levels, and different color interfaces correspond to different collision risk levels, so that the driver can see at a glance, so that the driver can intuitively and quickly prioritize emergencies, thereby improving driving safety.

[0043] The operation process of this embodiment is as follows:

[0044] (1) Classify and list several situations where the whistle needs to be blown:

[0045] like Figure 2 As shown in the figure, combined with common sense, four typical scenarios where horns are needed on highways or in urban-rural fringe areas are summarized, namely: Figure 2 As shown in (a), when overtaking, there is a car in front of you on the left, and you honk to warn it to prevent it from changing lanes suddenly; Figure 2As shown in (b), when overtaking, there is a car in front of you on the right, and you honk to warn it to prevent it from changing lanes suddenly; Figure 2 As shown in (c), when overtaking, there are cars on the left and right front, and the horn is honked to warn them to prevent them from changing lanes suddenly; Figure 2 As shown in (d), there are cars in front, on the left and on the right, and the horn is honked to urge them to move forward.

[0046] (2) Triggering of AC device:

[0047] The main controller on the vehicle is connected to the speed sensor on the vehicle and is used to control the power on and off of the AC device according to the vehicle speed detected by the speed sensor. The main controller and the electronic switch can be powered by the vehicle's own power supply. When the vehicle is turned on with a key, the vehicle's own power supply supplies power to the main controller and the electronic switch. At this time, if the vehicle is not moving and the speed sensor detects that the speed is 0, the main controller controls the electronic switch to be disconnected and the AC device does not work. If the vehicle is moving and the speed sensor detects that the speed is not 0, the main controller controls the electronic switch to be closed and the AC device is powered on. Therefore, when the vehicle starts to move, the AC device is automatically triggered to start working without the need for the driver to manually trigger it. The electronic switch of this embodiment can be a relay.

[0048] The AC device of this embodiment uses T as its working cycle (T will change with the vehicle speed, the faster the vehicle speed, the smaller T). Each working cycle can be roughly divided into four processes: sampling, calculation, signal sending and signal receiving. When the signal sending process in a working cycle is completed, the sampling process of the next working cycle is entered, and so on until the vehicle is turned off.

[0049] (3) Image acquisition by setting up a front depth camera:

[0050] like Figure 3 As shown, the depth camera collects a depth image of the road condition in front within its visual range, that is, collects a depth image in front of the vehicle and sends the depth image to the DSP module.

[0051] As mentioned above, the target vehicle is often located in front of the vehicle (including the front, left front, and right front). It is advisable to set a depth camera in front of the vehicle to collect images of the target vehicle group facing the front signal, that is, including potential target vehicles. The collected depth image includes not only the RGB value of each pixel, but also the depth information of each pixel. There are three options for the placement of the depth camera. The first is to place a depth camera on the outside of the front windshield. The advantage of this solution is simple layout and easy implementation; the second is to place a depth camera on the front side of the left rearview mirror and the front side of the right rearview mirror. The advantage is a large viewing angle and a high safety factor; the third is to place a depth camera on the front side of the rearview mirror inside the car. This solution is simple to implement.

[0052] (4) DSP module processes depth images:

[0053] like Figure 3 As shown, the DSP module obtains the license plate number string information and road width information of the potential target vehicle group from the depth image based on the license plate number recognition system; the DSP module obtains the position information of each target vehicle relative to the signal source vehicle from the depth image based on the principle of the depth camera.

[0054] After calculating the state parameters of the potential target vehicle group (i.e., the position of the target vehicle relative to the signal source vehicle and other information) and the road width based on the depth image, the target vehicles can be screened accordingly. Specifically, the DSP module sets a reasonable lateral distance threshold based on the road width, and records the target vehicles whose lateral distance to the vehicle is less than the lateral distance threshold as real target vehicles; and records the target vehicles whose lateral distance to the vehicle is greater than or equal to the lateral distance threshold as non-real target vehicles, i.e., irrelevant vehicles. Figure 4 As shown in the figure, the elimination principle of non-real target vehicles is:

[0055] X ≥ 2D;

[0056] Where X is the lateral distance between the potential target vehicle and the vehicle (i.e., the signal source vehicle), D is the road width, and 2D is the lateral distance threshold.

[0057] In this way, non-real target vehicles can be preliminarily eliminated from the potential target vehicle group, and the real target vehicles in the potential target vehicle group can be screened out, thereby narrowing the signal's facing range to a certain extent and achieving more accurate signal delivery.

[0058] Collision risk classification:

[0059] The DSP module calculates the collision time of the real target vehicle relative to the own vehicle based on the depth image, that is, the relative collision time TTC, and divides the collision risk level according to the relative collision time TTC.

[0060] Since drivers need to be fully focused when driving, the priority of avoiding multiple vehicles coming from different lanes behind will inevitably distract the driver's attention and be detrimental to driving safety. Therefore, it is advisable to divide the warning levels (collision risk levels) so that drivers can intuitively and quickly prioritize emergencies, thereby improving driving safety.

[0061] One idea for classifying the collision risk level of vehicles coming from behind is to calculate the relative collision time TTC based on the longitudinal speed and longitudinal distance of the actual target vehicle using a uniform motion model. Based on this, combined with the driver's reaction time and signal delay, a reasonable collision time threshold is set to classify the collision risk level of target vehicles with collision risks.

[0062] The relative collision time TTC refers to the time required for the vehicle to hit the vehicle in front at the current speed. The calculation of TTC is mainly based on the longitudinal speed, longitudinal distance and other state parameters of the real target vehicle relative to the vehicle (i.e. the signal source vehicle). The smaller the relative collision time TTC, the higher the danger level; the larger the relative collision time TTC, the lower the danger level. An idea of ​​calculating TTC using the depth image information captured by the depth camera and the strategy for dividing the danger level are as follows:

[0063] The frame numbers of the two depth images captured by the depth camera are recorded as N2 and N1 respectively, and the sampling frequency is f. The time interval Δt between the two depth images is:

[0064]

[0065] The longitudinal distances of the real target vehicle (i.e., the front vehicle) corresponding to the two depth images above relative to the host vehicle (i.e., the signal source vehicle) are d1 and d2 respectively. Then the longitudinal relative speed of the real target vehicle relative to the host vehicle is:

[0066]

[0067] At the current moment, the longitudinal distance of the vehicle relative to the real target vehicle is Y, and the relative collision time TTC is:

[0068]

[0069] Right now:

[0070]

[0071] According to the calculated relative collision time TTC, combined with the driver reaction time obtained from the survey (approximately 0.6s to 0.8s), the relative collision time threshold range is set, and the calculated relative collision time TTC is compared and analyzed with the relative collision time threshold range, so that the collision warning level (i.e., collision risk level) can be reasonably divided according to the danger level classification strategy shown in Table 1:

[0072] Table 1: Hazard classification strategy;

[0073]

[0074] Note: The first level of collision risk level is higher than the second level of collision risk level, the second level of collision risk level is higher than the third level of collision risk level, and the third level of collision risk level is higher than the fourth level of collision risk level.

[0075] Use the license plate string to locate the signal target and source: the license plate string of the real target vehicle is used as the address identifier of the signal target to achieve accurate signal delivery; the DSP module can obtain the following through coordinate transformation based on the position information of the real target vehicle relative to the vehicle: Figure 5 The position information of the host vehicle relative to the real target vehicle shown includes the longitudinal distance y of the host vehicle relative to the real target vehicle, the lateral distance x of the host vehicle relative to the real target vehicle, and the orientation information (ie, the angle θ) of the host vehicle relative to the real target vehicle.

[0076] The DSP module determines whether there is a collision risk based on the collision risk level, that is, whether the collision risk level belongs to level 1, 2 or 3. If so, it means that there is a certain collision risk and the signal needs to be encoded and broadcast to issue a warning signal.

[0077] Based on the Unicode encoding method, the DSP module integrates and encodes the license plate number string information of the real target vehicle, the license plate number string information of the vehicle itself, the position information of the vehicle relative to the real target vehicle, and the collision risk level information in sequence; the encoded data structure is: "license plate number string of the real target vehicle" + "license plate number string of the vehicle itself" + "position information of the vehicle relative to the real target vehicle" + "collision risk level" + "timestamp".

[0078] (5) Broadcasting of signals:

[0079] Using the DSRC protocol, the vehicle network module broadcasts the information encoded and integrated by the DSP module through the vehicle self-organizing network within a certain range in the form of "header" + "content". The "header" describes the license plate number of the vehicle (i.e. the real target vehicle) to which the signal is directed and the license plate number of the vehicle itself (corresponding to the first two items of the encoded data structure respectively); the "content" describes the position information of the vehicle relative to the real target vehicle and the collision risk level (corresponding to the last three items of the data structure respectively).

[0080] (6) Signal reception and decoding:

[0081] The vehicle network modules in the vehicles within the broadcast range receive the signal and send it to the DSP module in the vehicle; the DSP module decodes the signal based on the Unicode encoding table and compares the license plate number string of its own vehicle with the first item in the decoded information. If the comparison result is the same, the decoded information passes the verification, that is, the identity authentication is passed;

[0082] (7) Sound warning and information display:

[0083] The buzzer in the sound alarm module installed on the verified vehicle emits a sound warning at a frequency corresponding to the highest collision risk level in all the verified information. At the same time, the LCD display screen in the LCD display module installed on the verified vehicle displays the license plate number and position of all signal source vehicles in an interface according to the color corresponding to the collision risk level in the verified information. Different colors are used to inform the driver of the position and collision risk level of all signal source vehicles, so that the driver can intuitively and quickly prioritize the emergency situation, thereby improving driving safety. When a real target vehicle receives multiple verified information, that is, when multiple vehicles broadcast signals to the real target vehicle, the LCD display screen in the LCD display module of the real target vehicle uses different colors to represent the license plate number and position of different signal source vehicles (the specific color is set according to the collision risk level in each verified information), and the buzzer warns the frequency corresponding to the collision risk level with the highest collision risk in all the verified information. According to the buzzer frequency and LCD display color setting scheme shown in Table 2, the higher the danger level of the signal, the higher the sound frequency of the trigger buzzer; the LCD display screen displays different colors for signals of different danger levels. A buzzer frequency and LCD display color setting strategy is shown in Table 2. This can intuitively and efficiently inform the driver of the danger level of the warning signal.

[0084] Table 2: Buzzer frequency and LCD interface color setting scheme;

[0085] Collision risk level or warning level Level 4 Level 3 Level 2 Level 1 Buzzer frequency 0 2kHz 3kHz 5kHz LCD display interface color blue green orange color red

[0086] The higher the risk of collision, the higher the frequency of the buzzer, the more urgent the warning sound, and the more red the color of the warning interface. This setting strategy can fully demonstrate the collision risk level to the driver.

[0087] The protection scope of the present invention includes but is not limited to the above embodiments. The protection scope of the present invention shall be based on the claims. Any replacement, deformation, and improvement of the technology that can be easily thought of by technicians in this field shall fall within the protection scope of the present invention.

Claims

1. A green driver communication system based on machine vision and vehicle network, characterized by: The communication device includes a depth camera for collecting depth images, a DSP module for processing the depth images, an in-vehicle network module for signal broadcasting and signal reception, a sound alarm module for being installed in the vehicle and alerting the driver that a vehicle is approaching from behind, and an LCD display module for displaying the position of the vehicle from which the signal is coming, wherein the depth camera is connected to the DSP module, and the DSP module is connected to the in-vehicle network module, the sound alarm module, and the LCD display module; The depth camera is used to periodically collect depth image information of the target vehicle; In each working cycle, when the depth camera collects the depth image information of the target vehicle, the DSP module is used to obtain the license plate number string of the target vehicle from the depth image information based on the license plate number recognition system; The DSP module is used to obtain the position information of the target vehicle relative to the vehicle from the depth image information based on the principle of the depth camera, and the position information includes the lateral distance of the target vehicle relative to the vehicle, the longitudinal distance of the target vehicle relative to the vehicle, and the direction of the target vehicle relative to the vehicle; the DSP module is used to calculate the longitudinal speed of the target vehicle relative to the vehicle based on the position information obtained from the two frames of depth images; the DSP module is used to calculate the road width based on the collected depth image information; The vehicle network module is used to broadcast the encoded information in the DSRC protocol. All vehicles within the broadcast range can receive the information through the vehicle network module installed on their own vehicles. The DSP module installed on the vehicle that receives the information is used to decode the received information based on the Unicode encoding method. The DSP module is used to compare and verify the license plate number string of the own vehicle with the first item in the decoded information. When the comparison and verification result is the same, the decoded information passes the verification; if there is only one piece of information that passes the verification, the buzzer in the sound alarm module installed on the vehicle that passes the verification is used to emit a buzzer at a frequency corresponding to the collision risk level in the information that passes the verification. Sound warning, the LCD display module installed on the vehicle that has passed the verification is used to display the position of the signal source vehicle according to an interface with a color corresponding to the collision risk level in the information that has passed the verification; if there are multiple pieces of information that have passed the verification, the DSP module installed on the vehicle that has passed the verification is used to select the highest collision risk level from all the information that has passed the verification, and the buzzer in the sound alarm module installed on the vehicle that has passed the verification is used to issue a sound warning at a frequency corresponding to the highest collision risk level. At the same time, the LCD display module installed on the vehicle that has passed the verification is used to display the positions of all signal source vehicles according to an interface with a color corresponding to the collision risk level in the decoded information.

2. The green driver communication system based on machine vision and vehicle network according to claim 1 is characterized in that: It also includes a power module, an electronic switch and a main controller. The power module is connected to the AC device through the electronic switch, the main controller is connected to the vehicle speed sensor on the vehicle, and the main controller is connected to the electronic switch. The main controller is used to control the on and off of the electronic switch according to the vehicle speed detected by the speed sensor, thereby controlling the power on and off of the AC device.

3. The green driver communication system based on machine vision and vehicle network according to claim 2 is characterized in that: The depth camera is arranged at the front windshield of the vehicle, at the left and right rearview mirrors, or at the front side of the rearview mirror inside the vehicle.

4. The green driver communication system based on machine vision and vehicle network according to claim 1 is characterized in that: The DSP module is used to screen out the real target vehicle based on the road width and the position information of each target vehicle in the target vehicle group relative to the vehicle, and calculate the relative collision time between the real target vehicle and the vehicle based on the position information and longitudinal speed of the real target vehicle relative to the vehicle, and then classify the collision risk level according to the relative collision time.

5. The green driver communication system based on machine vision and vehicle network according to claim 4 is characterized in that: The DSP module is used to determine whether there is a collision risk based on the collision risk level. If so, the DSP module integrates and encodes the license plate number string information of the real target vehicle, the license plate number string information of the vehicle, the position information of the vehicle relative to the real target vehicle, and the collision risk level information in sequence based on the Unicode encoding method.

Images