TWI761181B - Vehicle driving image recognition system - Google Patents

Abstract

The invention provides a vehicle driving image recognition system, which includes an acceleration sensing module and a camera device. The acceleration sensing module senses the acceleration value of the vehicle. The camera device includes a camera module, an image recognition module, a signal processing module and an image transmission module. The camera module captures the driving state of the driving in the vehicle. The image recognition module recognizes the image captured by the camera module. The signal processing module is electrically connected to the acceleration sensing module, the camera module and the image recognition module. The image transmission module is electrically connected to the image recognition module to transmit the driving state of driving. When the acceleration sensing module senses that the acceleration value of the vehicle exceeds the preset value, a trigger signal is generated to the signal processing module, and the signal processing module generates a trigger signal. A control signal is generated to the image recognition module, and the image transmission module sends the driving status of the driver to the cloud server as a reference for judging whether the driver is responsible for the accident.

Description

Vehicle driving image recognition system

The present invention relates to an image recognition system, in particular to a vehicle driving image recognition system.

Vehicles with automatic driving function are the development trend of today's automobile industry. Through the function of automatic driving, the spirit of driving in manual driving can be saved. However, the automatic driving function is not provided to make the driver not need to control the vehicle and pay attention to the road conditions at all, but to facilitate driving. The automatic driving function can increase the convenience of vehicle control in terms of vehicle control.

As mentioned above, if the driver abuses the convenience of automatic driving excessively, pays attention to the road conditions and does not focus on the control of the vehicle, it is difficult to clarify that it is caused by the misjudgment of the automatic driving of the vehicle itself, or It is caused by improper operation of driving, and it is impossible to clarify the attribution of responsibility.

In addition, even if the driver does not turn on the automatic driving function, in the event of a car accident, the insurance company must still make claims based on the mental state of the driver, the state of vehicle control, and the situation of the car accident.

However, regardless of the above situation, under the principle of not self-incriminating, the driver may not truthfully inform the insurance company of his mental state and vehicle control state at that time, or if the driver wants to prove that his mental state is good and the vehicle is in good condition. All of them are under normal control, which will make it difficult for insurance companies to make objective claims based on the inability to record and identify the mental state of the real driving and the state of vehicle control.

Accordingly, how to provide a vehicle driving image recognition system to solve the above problems has become an urgent research topic.

In view of the above problems, the present invention provides a vehicle driving image recognition system, which includes an acceleration sensing module and a camera device. The acceleration sensing module senses the acceleration value of the vehicle. The camera device is arranged in the vehicle and includes a camera module, an image recognition module, a signal processing module and an image transmission module. The camera module captures the driving state of the driving in the vehicle. The image recognition module recognizes the image captured by the camera module. The signal processing module is electrically connected to the acceleration sensing module, the camera module and the image recognition module. The image transmission module is electrically connected to the image recognition module, wherein when the acceleration sensing module senses that the acceleration value of the vehicle exceeds a preset value, the acceleration sensing module generates a trigger signal to the signal processing module, and the signal processing module A control signal is generated to the camera module, and the captured image is captured by the camera module and then sent to the image recognition module for image analysis and recognition, and then the image transmission module transmits the recognized driving status image.

Continuing from the above, the vehicle driving image recognition system of the present invention records and recognizes various driving states of the vehicle, so as to determine whether the responsibility for the accident is caused by the driver’s own negligence (such as the above various driving states when a car accident occurs). ), or caused by the wrong decision of the automatic driving system (such as the image recognition error and hitting an obstacle, pedestrian or other vehicle), clarifying that the responsibility for the accident lies on one's own vehicle, other vehicles or other passers-by, and further serves as insurance Reference basis for company claims.

Please refer to FIG. 1A and FIG. 1B , which are schematic block diagrams and usage scenarios of the first embodiment of the vehicle driving image recognition system of the present invention. The vehicle driving image recognition system 1 includes an acceleration sensing module 11 and a camera device 12 . The acceleration sensing module 11 senses the acceleration value of the vehicle. The camera device 12 is disposed in the vehicle, and includes a first camera module 121 , a signal processing module 122 , an image recognition module 128 and an image transmission module 123 . The first camera module 121 captures the driving state of the driving in the vehicle. The image recognition module 128 recognizes the driving state of the driver captured by the first camera module 121 . The signal processing module 122 is electrically connected to the acceleration sensing module 11 , the first camera module 121 and the image recognition module 128 . The image transmission module 123 is electrically connected to the image recognition module 128. When the acceleration sensing module 11 senses that the acceleration value of the vehicle exceeds the first preset value, the acceleration sensing module 11 generates a first trigger signal to the signal processing module. In the group 122, the signal processing module 122 generates the first control signal to the image recognition module 128 to analyze and recognize the image, and then the image transmission module 123 sends the driving state image of the recognized driving.

In an embodiment of the present invention, the camera device 12 can be installed on the windshield in the vehicle, or on the vehicle platform under the windshield. The installation position of the camera device 12 is not limited in the present invention, but can be installed In any location where you can clearly photograph the driving state. In addition, the first camera module 121 of the camera device 12 has a shooting lens with a 360-degree panoramic view, which can shoot front, rear, left, and right images inside the vehicle and front, rear, left, and right images outside the vehicle. Furthermore, the camera device 12 has an orientation adjustment element 124 , which can adjust and rotate the angle of the first camera module 121 on the camera device 12 to further change the position where the first camera module 121 shoots. In addition, it should be noted that the first camera module 121 in the camera device 12 starts the function of recording and shooting images when the vehicle is started, that is, it is a continuous shooting image, not when the vehicle collides Just take the action.

In the embodiment of the present invention, the image recognition module 128 (including the artificial intelligence image recognition software and the database) can be directly installed in the camera device 12 of the vehicle driving image recognition system 1, and perform images on the local end (driving the vehicle). analysis, identification and comparison, and send the analysis, identification and comparison results to the cloud server C for storage. Alternatively, in another embodiment of the present invention, the image recognition module 128 may be disposed in the cloud server C, which is not limited in the present invention. In other words, when the image recognition module 128 is disposed in the camera device 12 , the camera device 12 performs image recognition on the local end after shooting and capturing images through the first camera module 121 . When the image recognition module 128 is installed in the cloud server C, the camera device 12 shoots and captures the image through the first camera module 121, and then transmits the image to the cloud server C (remote) through the image transmission module 123 for processing. The action of image recognition. The recognition process takes the driver's eye state as an example. If the image recognition module 128 captures several images (frames) shot and captured by the first camera module 121 for a period of time, it is recognized that the driver has eyes closed. , the signal processing module 122 can determine that the driving is in a state of dangerous driving, and when a car accident occurs, the driver of the vehicle can be held responsible for the accident according to the dangerous state.

Please refer to FIG. 2A and FIG. 2B , which are schematic block diagrams and usage scenarios of the second embodiment of the vehicle driving image recognition system of the present invention. In this embodiment, the camera device 12 of the vehicle driving image recognition system 1 further includes a second camera module 125, which is arranged in the vehicle corresponding to the position of the brake and the accelerator of the vehicle, or is arranged in the vehicle corresponding to the position of the driver's feet. The signal processing module 122 and the image recognition module 128 are sexually connected, and the step status of the driver in the vehicle is captured, including whether the driver's left foot is on the brake or the accelerator position and the status of the entire leg (whether there is a foot-swinging behavior). ), and photographing the facial expressions and body movements of the driver, wherein when the acceleration sensing module 11 senses that the acceleration value of the vehicle exceeds a first preset value, the acceleration sensing module 11 generates a first trigger signal to the signal processing module In the group 122, the signal processing module 122 generates the first control signal to the image recognition module 128, and after the image recognition module 128 recognizes the driving state image of the driving captured by the second camera module 125, the image transmission module 123 Send the identified driving footstep status to the cloud server C for storage. Similarly, the second camera module 125 in the camera device 12 starts the function of recording and capturing images when the vehicle is started, that is, it is an image that is continuously captured, not captured when the vehicle collides. action.

Please refer to FIG. 3 , which is a schematic diagram of the vehicle of the present invention colliding with an obstacle. Continuing from the above, as shown in FIG. 3 , when the vehicle collides, the acceleration value of the vehicle sensed by the acceleration sensing module 11 will instantly exceed the first preset value. Therefore, the image transmission module 123 uses the Internet The driving status image and the driving footstep status image are sent to the cloud server C for identification, analysis and storage, or the image identification and analysis are performed by the local image identification module. Cloud server C includes Software as a Service, Platform as a Service, and Infrastructure as a Service. All kinds of services are available through Artificial Intelligence. ; AI) image recognition technology, image recognition software and database for training, identification, analysis and comparison of the transmitted driving status images. Dozing off with eyes closed, blinking continuously), mouth closed status (whether you keep opening your mouth due to fatigue), limb status (whether you bend over to pick up items, whether you are wearing seat belts), head position status (whether you turn your head? , bowing head), gesture status (whether holding a cigarette and smoking, whether holding a smart device), footstep status (including accelerator and brake control status), sound status (such as whether driving is distracted by chatting with passengers), steering wheel/direction light control Status (including whether the driver's hands are holding the steering wheel and pressing the direction light), and according to each status, the result of the judgment of the responsibility of the car accident corresponding to the driving is generated.

Please refer to FIG. 4 , which is a block diagram of a third embodiment of the vehicle driving image recognition system of the present invention. In the above embodiment, in addition to uploading the image data to the cloud server C, in this embodiment, the camera device 12 further includes a storage module 126. The storage module 126 is electrically connected to the first camera module 121 and the second camera module 121. The camera module 125 , the image recognition module 128 and the image transmission module 123 store the driving state and the step state of the driver. In other words, the images captured by the first camera module 121 and the second camera module 125 can be stored locally or in the cloud, which is not limited in the present invention.

Furthermore, the vehicle driving image recognition system of the present invention can be applied not only to a vehicle with an automatic driving function, but also to a general manual driving vehicle, so as to clarify the responsibility of the accident in the event of a car accident.

Please refer to FIG. 5 , which is a block diagram of a fourth embodiment of the vehicle driving image recognition system of the present invention. In this embodiment, the vehicle driving image recognition system 1 further includes an alcohol concentration detector 127, which is electrically connected to the signal processing module 122, and is disposed in the vehicle and in front of the driver's seat to detect the alcohol concentration value of driving near the steering wheel. . When the alcohol concentration detector 127 senses that the alcohol concentration value of driving exceeds the second preset value for a period of time, the alcohol concentration detector 127 can determine that the driver is drinking, and the alcohol concentration detector 127 generates a first Two trigger signals are sent to the signal processing module 122, and the second control signal is generated by the signal processing module 122 to the image recognition module 128. After the image recognition module 128 analyzes and recognizes the driving image, the image transmission module 123 sends the driving image The driving status is sent to the cloud server C.

In an embodiment of the present invention, the acceleration sensing module 11 (G-Sensor) is disposed on the vehicle. In another embodiment of the present invention, the acceleration sensing module 11 is disposed on the camera device 12 . The acceleration sensing module 11 can respectively set a first preset value for the acceleration value generated when the vehicle collides with a pedestrian, an obstacle or collides with other vehicles, and when the acceleration value generated by the collision exceeds the first preset value A first trigger signal is generated. In other words, when the acceleration value exceeds the first preset value, it means that the driving vehicle collides. At this time, the acceleration sensing module 11 generates a trigger signal to the signal processing module 122, and the signal processing module 122 generates the first trigger signal. The control signal and/or the second control signal are sent to the image recognition module 128, and after the image recognition module 128 analyzes and recognizes the images captured by the first camera module 121 and the second camera module 125, it is sent by the image transmission module 123 The driving status of the driver is sent to the cloud server C.

To sum up, the vehicle driving image recognition system of the present invention records and recognizes various driving states of the vehicle, so as to determine whether the responsibility for the accident is caused by the driver's own negligence (such as the above various driving states when a car accident occurs). ), caused by the external environment (for example, the image of the ground obstacle cannot be clearly identified, the road conditions cannot be clearly identified due to the dark sky), or caused by the wrong decision of the automatic driving system (such as the image recognition error and hitting the obstacle objects, pedestrians or other vehicles), and clarify that the responsibility for the accident lies with one’s own vehicle, another’s vehicle, or other passers-by, and further serve as a reference for insurance companies to settle claims.

1: Vehicle driving image recognition system 11: Acceleration sensing module 12: Camera device 121: The first camera module 122: Signal processing module 123: Image transmission module 124: Azimuth adjustment element 125: Second camera module 126: Storage Module 127: Alcohol concentration detector 128: Image recognition module C: cloud server

FIG. 1A and FIG. 1B are a block schematic diagram and a usage situation diagram of a first embodiment of a vehicle driving image recognition system according to the present invention; 2A and 2B are a schematic block diagram and a usage situation diagram of a second embodiment of the vehicle driving image recognition system of the present invention; 3 is a schematic diagram of the vehicle of the present invention colliding with an obstacle; FIG. 4 is a block diagram of a third embodiment of a vehicle driving image recognition system according to the present invention; and FIG. 5 is a block diagram illustrating a fourth embodiment of a vehicle driving image recognition system according to the present invention.

1: Vehicle driving image recognition system

11: Acceleration sensing module

12: Camera device

121: The first camera module

128: Image recognition module

122: Signal processing module

123: Image transmission module

C: cloud server

Claims (11)

A vehicle driving image recognition system, comprising: an acceleration sensing module for sensing an acceleration value of a vehicle; and a camera device disposed in the vehicle, comprising: a first camera module for photographing images in the vehicle a driving status; an image recognition module electrically connected to the first camera module to identify the driving status of the driver; a signal processing module electrically connected to the acceleration sensing module, the first camera module a camera module and the image recognition module; an image transmission module, electrically connected to the image recognition module, to transmit the driving state of the driver after recognition; and a second camera module, corresponding to a brake and a brake of the vehicle The accelerator position is set in the vehicle, is electrically connected to the signal processing module and the image recognition module, and captures a step state of the driver in the vehicle, wherein when the acceleration sensing module senses the When the acceleration value exceeds the first preset value, the signal processing module generates the first control signal to the image recognition module, and the image recognition module recognizes the step state of the driver, and then transmits the image through the image The module sends the step state of the driving; wherein when the acceleration sensing module senses that the acceleration value of the vehicle exceeds a first preset value, the acceleration sensing module generates a first trigger signal to the signal a processing module, the signal processing module generates a first control signal to the image recognition module, after the image recognition module recognizes the driving state of the driver, the image transmission module sends the driver's driving status. The vehicle driving image recognition system according to claim 1, wherein the acceleration sensing module is disposed on the vehicle. The vehicle driving image recognition system according to claim 1, wherein the acceleration sensing module is disposed on the camera device. The vehicle driving image recognition system according to claim 1, wherein the first preset value includes the acceleration value generated when the vehicle collides with a pedestrian, collides with an obstacle or collides with a vehicle. The vehicle driving image recognition system according to claim 1, wherein the driving state of the driver includes a mental state, an eye state, a mouth closed state, a limb state, a head position state, a gesture state, a step state status, a sound status, and a steering wheel/direction light control status. The vehicle driving image recognition system according to claim 1, wherein the camera device further comprises a storage module, which is electrically connected to the first camera module and the second camera module, and stores the driving state and the driving state of the driver. The step state. The vehicle driving image recognition system according to claim 5, wherein the image recognition module recognizes whether the mental state of the driver is a state where the temperature of the face continues to rise, and whether the eye state is dozing off with closed eyes or continuous blinking state, identify whether the closed mouth state is the state of continuously opening the mouth, identify whether the limb state is the state of bending over or not wearing a seat belt, identify whether the head position state is the state of turning the head or bowing the head, identifying the state of the body Whether the gesture state is holding a cigarette, smoking or holding a smart device, identifying whether the step state is placed in the accelerator or braking position, identifying whether the sound state is chatting with passengers, identifying whether the steering wheel/direction light control state is the driving Holding both hands on the steering wheel and pressing down the direction lights, and generating a recognition result to the signal processing module according to each state, the signal processing module correspondingly generates the driver's responsibility for a car accident according to the recognition result critical result. The vehicle driving image recognition system as claimed in claim 1, further comprising an alcohol concentration detector, electrically connected to the signal processing module, disposed in the vehicle and detecting alcohol in the driving concentration value, wherein when the acceleration sensing module senses that the acceleration value of the vehicle exceeds the first preset value, and when the alcohol concentration detector senses that the alcohol concentration value of the driving exceeds a second preset value When the value is set, the signal processing module generates a second control signal to the image transmission module to send the driving state of the driving. The vehicle driving image recognition system according to claim 1, wherein the vehicle is an automatic driving vehicle. The vehicle driving image recognition system according to claim 1, wherein the first camera module has a 360-degree surround view shooting lens. The vehicle driving image recognition system as claimed in claim 1, wherein the driving state is transmitted to a cloud server through an Internet, and the cloud server analyzes and recognizes the driving state of the driver, and according to the recognized The driving state of the driving corresponds to the judgment result of the responsibility for causing a car accident.

Images