An intelligent electronic rearview mirror system based on human eye attention and implementation method
The display range of the electronic rearview mirror and central LCD is automatically adjusted through the human eye attention detection unit, which solves the problem of unclear vision and inclement of driver attention changes in traditional optical rearview mirrors in bad weather, improving driving safety and environmental visibility.
Patent Information
- Application Number
- CN202111399057.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-11-19
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2041-11-19
AI Technical Summary
Traditional optical rearview mirrors are difficult to provide a clear field of view in bad weather and cannot be automatically adjusted to cope with drivers' attention changes, resulting in visual blind spots and driving safety issues.
The human eye attention detection unit is used to detect the driver's head attitude and line of sight direction through a monocular camera, and combine it with MTCNN and head attitude network to adjust the display range of the electronic rearview mirror and central control LCD, providing automatic adjustment and dangerous target detection functions. .
Improves driver visibility into the surrounding environment, reduces operation volume, enhances driving safety, and maintains a clear field of view in severe weather, providing real-time detection and reminders of dangerous targets.
Smart Images

Figure CN113978366B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the fields of intelligent driving, assisted driving, deep learning, intelligent cockpit technology, and more specifically to an intelligent electronic rearview mirror system based on human eye attention and an implementation method thereof. Background Art
[0002] Traditional vehicle driving mainly relies on the optical rearview mirror protruding from the side of the vehicle body to observe the situation around the vehicle, thereby adjusting the vehicle's posture and completing operations such as changing lanes, overtaking, and reversing. In traditional optical rearview mirrors, the curvature, shape, and size of the reflective mirror surface are fixed. At most, the curvature of the lens is increased on the outer edge, sacrificing the integrity of the picture in exchange for a wider field of view. This is what we often call a hyperbolic rearview mirror. Incorrect manual adjustment of the direction of the optical rearview mirror can easily produce a large visual blind spot, affecting driving safety. At the same time, the optical rearview mirror is installed on the outside of the car, so it will be affected by the environment such as rainy days or wet ground in summer. In bad weather, it is difficult for the driver to clearly understand the situation on the left, right, and rear of the car through the rearview mirror outside the car, affecting driving safety. The optical rearview mirror protrudes from the body, has great wind resistance, and produces noise when driving at high speed. At the same time, the traditional optical rearview mirror cannot provide danger warnings for the driver. When the driver changes lanes, turns, or when dangerous targets approach around, if the driver does not pay attention to the rearview mirror in time, danger is likely to occur.
[0003] After searching, the reference document 1, the application publication number is CN111645521A, a control method, device, electronic device and storage medium for a smart rearview mirror, which relates to the field of vehicle networking technology. The specific implementation scheme is: obtaining the interface display information of the smart rearview mirror; sending the interface display information to at least one external device connected to the smart rearview mirror; responding to the user operation information from at least one external device, controlling the smart rearview mirror to perform the corresponding operation, and the user operation information is used to describe the user's operation on at least one external device. The technical solution disclosed in this application can realize that the user controls the interface of the smart rearview mirror projection on the external device of the projection screen, which is easy to operate and can ensure the user's driving safety. The invention realizes that the external device displays the smart rearview mirror information and sets the smart rearview mirror through the external display device, but the driver still needs to perform manual operation, which does not enhance driving safety. The present invention adopts an attention detection unit to obtain the driver's attention direction, and directly adjusts the left rear electronic rearview mirror and the central control LCD display screen, which reduces the driver's operation and ensures driving safety.
[0004] Comparative Document 2, application publication number CN111625094A, an interactive method, device, electronic device and storage medium for a smart rearview mirror, relating to the field of artificial intelligence. In this application, the action interaction with the user can be realized based on the front camera on the smart rearview mirror. On the premise of increasing the interaction method with the user, it can also solve the problem that the existing voice interaction cannot recognize the user's voice in specific scenarios. In this application, it is also possible to control the smart rearview mirror to interact with the user according to the current use status of the smart rearview mirror, so that the smart rearview mirror can interact with the user in a convenient interactive way according to the current use status of the smart rearview mirror, thereby improving the interaction success rate and thus improving the user experience. This patent realizes the interaction between the smart rearview mirror and the driver's actions and sounds, which can meet the driver's multimedia needs such as navigation and telephone, and improve the user's driving experience. However, this patent does not pay more attention to the driver's safety needs while driving. This patent realizes display adjustment under different attention levels through human eye attention detection units, target detection units, etc., so that the driver can observe a larger area, and at the same time provides a dangerous object detection function. When the driver is changing lanes, turning, etc., the central control LCD and the left and right electronic rearview mirror screens can display surrounding dangerous objects to remind the driver to pay attention, thereby improving driving safety. Summary of the invention
[0005] The present invention aims to solve the above problems of the prior art. An intelligent electronic rearview mirror system based on human eye attention and an implementation method are proposed. The technical solution of the present invention is as follows:
[0006] An intelligent electronic rearview mirror system based on human eye attention, comprising:
[0007] An object detection unit, used to detect the type and position of objects around the vehicle to obtain the driving status of the vehicle;
[0008] The display unit uses the electronic rearview mirror screen and the central control LCD as display devices to display the images collected by the camera on the electronic rearview mirror screen and the central control LCD; when the driver performs operations such as changing lanes and turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle;
[0009] The human eye attention detection unit uses a monocular camera as a detection device to detect the driver's head posture, sight direction and attention direction; the human eye attention detection unit uses MTCNN to perform accurate face detection, obtains the driver's head posture through a head posture network, obtains the driver's sight direction using a sight detection network, and finally obtains the driver's attention direction by combining the head posture and sight direction;
[0010] The adjustment unit is used to adjust the field of view of the central control LCD display and the electronic rearview mirror screen to ensure that the display range of the central control LCD display and the electronic rearview mirror screen is consistent with the line of sight of human eyes.
[0011] Furthermore, the target detection unit uses a target detection network to detect and classify surrounding objects. The target detection network specifically includes the following sub-steps:
[0012] A target detection network based on YOLOv3 is constructed. By combining the KITTI public dataset and the self-sampling dataset, the vehicle runtime environment targets are divided into five categories: person, bike, car, truck, and bus. The target detection network is trained.
[0013] Furthermore, the display unit displays the image captured by the camera on the electronic rearview mirror screen and the central control LCD; at the same time, it detects targets in the vehicle's surrounding environment. When the driver performs operations such as changing lanes and turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle.
[0014] Furthermore, the eye attention detection unit uses MTCNN to perform accurate face detection, obtains the driver's head posture through the head posture network, obtains the driver's line of sight direction using the line of sight detection network, and finally obtains the driver's attention direction by combining the head posture and the line of sight direction; specifically includes:
[0015] The human eye attention detection network takes the driver image captured by the monocular camera as input and outputs the driver's attention direction;
[0016] The human eye attention detection network first uses the MTCNN multi-task convolutional neural network to extract the face of the driver image captured by the monocular camera to obtain the face area;
[0017] Then, the head posture estimation algorithm is used to detect the face area and obtain the Euler angles (yaw angle, pitch angle, roll angle) of the head posture. At the same time, the face area is input into the sight line detection neural network to obtain the driver's sight line direction;
[0018] Finally, the driver's line of sight direction is converted into Euler angles and combined with the Euler angles of the head posture for calibration, and finally the driver's attention direction is obtained.
[0019] Furthermore, the adjustment unit uses the driver's head posture and line of sight direction obtained by the human eye attention detection unit to continuously adjust the screen display content of the electronic rearview mirror on the corresponding side according to the driver's attention direction; at the same time, according to the driver's attention direction, the lane information on different sides is displayed on the central control LCD display screen.
[0020] A method for implementing the system according to any one of the above items, comprising the following steps:
[0021] Step 1: Build a human eye attention detection network. Place a monocular camera in front of the driver to take a photo of the driver's face and input it into the network to obtain the driver's head posture and line of sight direction.
[0022] Step 2: When the driver looks to the left, the display screen installed at the left A-pillar position displays the left side and rear view of the car captured by the camera on the left rearview mirror and adjusts the display angle from angle a to angle a1 to increase the display range of the left side and rear view of the car. The central control LCD screen displays the left lane scene, and the central control LCD keeps displaying after the line of sight returns to the right position;
[0023] Step 3, when the driver's line of sight is to the right, the display screen installed at the right A-pillar position displays the right side and rear side images of the car captured by the camera on the right rearview mirror and adjusts the display angle from angle b to angle b1 to increase the display range of the right side and rear side images of the car, and the central control LCD display screen displays the right lane scene. After the line of sight returns to the right side, the central control LCD display screen keeps displaying;
[0024] Step 4: Build a target detection network to detect targets in the vehicle's surrounding environment. When the driver performs operations such as changing lanes or turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays dangerous objects around the vehicle.
[0025] Furthermore, the human eye attention detection network in step 1 takes the driver image captured by the monocular camera as input and outputs the driver's attention direction;
[0026] The human eye attention detection network first uses the MTCNN multi-task convolutional neural network to extract the face of the driver image captured by the monocular camera to obtain the face area;
[0027] Then, the head posture estimation algorithm is used to detect the face area and obtain the Euler angles (yaw angle, pitch angle, roll angle) of the head posture. At the same time, the face area is input into the sight line detection neural network to obtain the driver's sight line direction;
[0028] Finally, the driver's line of sight direction is converted into Euler angles and combined with the Euler angles of the head posture for calibration, and finally the driver's attention direction is obtained.
[0029] Furthermore, the target detection network specifically includes the following sub-steps:
[0030] A target detection network based on YOLOv3 is constructed. By combining the KITTI public dataset and the self-sampling dataset, the vehicle runtime environment targets are divided into five categories: person, bike, car, truck, and bus. The target detection network is trained.
[0031] A storage medium stores a computer program, which implements any one of the implementation methods when executed by a processor.
[0032] The advantages and beneficial effects of the present invention are as follows:
[0033] The purpose of the present invention is to provide an intelligent electronic rearview mirror system based on human eye attention in order to overcome the defects of traditional optical rearview mirrors. The system automatically adjusts the display range of the electronic rearview mirrors on both sides according to the driver's line of sight, reduces the safety problems caused by blurred rearview mirrors due to the installation angle of the rearview mirrors and bad weather, and improves driving safety. At the same time, it provides a target detection function based on the target detection network. When the driver changes lanes, turns, etc., or dangerous targets approach around, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays dangerous objects around the vehicle.
[0034] Specifically, it is manifested in the following points:
[0035] (1) The present invention uses a monocular camera to measure the driver's head posture and line of sight direction under the intelligent driving and automobile intelligent cockpit technologies. When the driver enters the cab and sits down, the human eye attention detection network detects the driver's head height and line of sight direction. When the driver's attention is deflected, the display screen automatically adjusts the display range, so that the driver can clearly understand the road conditions in a larger range around the car, thereby improving driving safety.
[0036] (3) Since the display screen of the present invention is installed inside the car, it is not affected by rainy days or wet ground in summer. The driver can clearly understand the situation on the left side, right side and rear of the car through the display screen inside the car, thereby improving driving safety.
[0037] (4) The present invention provides a dangerous target detection function based on a target detection network. When a dangerous target is detected approaching, the left and right electronic rearview mirror screens and the central control LCD simultaneously display the marked dangerous object, thereby improving driving safety.
[0038] (5) The present invention uniquely uses a human eye attention detection network to detect the direction of the driver's attention, thereby adjusting the images of the left and right electronic rearview mirror screens and the central control LCD, realizing automatic adjustment of the electronic rearview mirror without the driver having to perform additional actions and command outputs, thereby minimizing the driver's operation workload and enhancing driving safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 It is a flow chart of an intelligent electronic rearview mirror system based on human eye attention according to a preferred embodiment of the present invention.
[0040] Figure 2 is a schematic diagram of human eye attention detection in the present invention;
[0041] Figure 3 This is a schematic diagram of the human eye attention detection network detecting the driver's face angle in the present invention;
[0042] Figure 4 It is a schematic diagram of the effect of adjusting the left side picture angle of the car based on the human eye attention intelligent electronic rearview mirror system in the present invention;
[0043] Figure 5 It is a schematic diagram of the effect of adjusting the angle of the right side of the car based on the human eye attention intelligent electronic rearview mirror system in the present invention;
[0044] Figure 6 It is a schematic diagram of the effect of a car changing lanes based on the intelligent electronic rearview mirror system of the present invention based on human eye attention. DETAILED DESCRIPTION
[0045] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.
[0046] The technical solution of the present invention to solve the above technical problems is:
[0047] The purpose of the present invention is to provide an intelligent electronic rearview mirror system based on human eye attention, which is used in the field of automobile intelligent driving and intelligent cockpit technology. By measuring the driver's head posture and line of sight direction through a monocular camera, the electronic rearview mirror target detection and display based on human eye attention is completed. This technical solution is proposed, such as Figure 1 As shown, the following steps are included:
[0048] Step 1, such as Figure 2 As shown in the figure, a human eye attention detection network is constructed to detect the driver's line of sight. A monocular camera is installed in front of the driver so that the camera can obtain the driver's face image, collect the driver's different head postures and different line of sight data, and combine the network data set to perform accurate face detection by using MTCNN, and obtain the driver's head posture through the head posture network. The driver's line of sight direction is obtained using the line of sight detection network, and finally the driver's attention direction is obtained by combining the head posture and line of sight direction.
[0049] Step 2: When the driver looks to the left, the display screen installed at the left A-pillar position displays the left and rear images of the car captured by the camera on the left rearview mirror and adjusts the display angle from angle a to angle a1 to increase the display range of the left and rear images of the car. The central control LCD display screen displays the left lane scene, and the central control LCD keeps displaying after the line of sight returns to the right.
[0050] Step 3: When the driver looks to the right, the display screen installed at the right A-pillar position displays the right side and rear side images of the car captured by the camera on the right rearview mirror and adjusts the display angle from angle b to angle b1 to increase the display range of the right side and rear side images of the car. The central control LCD screen displays the right lane scene, and the central control LCD keeps displaying after the driver's sight returns to the right side.
[0051] Step 4, such as Figure 6 As shown in the figure, a target detection network is constructed, and the network data set and actual road data are used for data annotation and training to obtain the target detection network. The target detection network is used to detect targets in the vehicle's surrounding environment. When the driver performs operations such as changing lanes or turning, or detects that danger is approaching, such as surrounding vehicles approaching, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays dangerous objects around the vehicle.
[0052] The system also includes:
[0053] An object detection unit, used to detect the type and position of objects around the vehicle to obtain the driving status of the vehicle;
[0054] The display unit uses the electronic rearview mirror screen and the central control LCD as display devices to display the images collected by the camera on the electronic rearview mirror screen and the central control LCD. When the driver changes lanes, turns, or dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle;
[0055] The human eye attention detection unit uses a monocular camera as a detection device to detect the driver's head posture, line of sight direction and attention direction.
[0056] The adjustment unit is used to adjust the field of view of the central control LCD display and the electronic rearview mirror screen to ensure that the display range of the central control LCD display and the electronic rearview mirror screen is consistent with the line of sight of human eyes.
[0057] Furthermore, the target detection unit uses a target detection network to detect and classify surrounding objects.
[0058] Furthermore, the display unit displays the image collected by the camera on the screen of the electronic rearview mirror and the central control LCD. At the same time, the target environment around the vehicle is detected. When the driver performs operations such as changing lanes or turning, or when a dangerous target approaches, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle.
[0059] Furthermore, the human eye attention detection unit is equipped with a monocular camera in front of the driver so that the camera can obtain the driver's facial image, collect different head postures and different line of sight data of the driver, and combine the network data set to perform accurate face detection by utilizing MTCNN, and obtain the driver's head posture through the head posture network, and obtain the driver's line of sight direction using the line of sight detection network, and finally obtain the driver's attention direction by combining the head posture and line of sight direction.
[0060] Furthermore, the adjustment unit uses the driver's head posture and sight direction obtained by the human eye attention detection unit to continuously adjust the screen display content according to the movement of the driver's sight line. At the same time, according to the driver's attention direction, the lane information on different sides is displayed on the central control LCD display screen.
[0061] Furthermore, the intelligent electronic rearview mirror system shown further includes:
[0062] Monocular camera: The monocular camera is installed in front of the driver. During normal driving observation behavior, the camera can effectively observe the driver's left and right eyes.
[0063] External body camera: There are two external body cameras, namely, a left camera and a right camera. The left camera is set at the position of the left rearview mirror of the car, and the right camera is set at the position of the right rearview mirror of the car.
[0064] Display screen: There are also two electronic rearview mirror display screens in the car, namely the left display screen and the right display screen. The left and right display screens are placed on the left and right A-pillars of the vehicle respectively. The left display mirror matches the left camera, and the right display mirror matches the right camera.
[0065] Central control LCD: The central control LCD display is the LCD screen that comes with the car's center console.
[0066] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0067] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0068] 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.
[0069] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.
Claims
1. An intelligent electronic rearview mirror system based on human eye attention, characterized in that: include: An object detection unit, used to detect the type and position of objects around the vehicle to obtain the driving status of the vehicle; The display unit uses the electronic rearview mirror screen and the central control LCD as display devices to display the images collected by the camera on the electronic rearview mirror screen and the central control LCD; when the driver performs operations such as changing lanes and turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle; The human eye attention detection unit uses a monocular camera as a detection device to detect the driver's head posture, line of sight and attention direction; The eye attention detection unit uses MTCNN to perform accurate face detection, obtains the driver's head posture through the head posture network, obtains the driver's line of sight direction using the line of sight detection network, and finally obtains the driver's attention direction by combining the head posture and line of sight direction; An adjustment unit is used to adjust the visual range of the central control LCD display and the electronic rearview mirror screen to ensure that the display range of the central control LCD display and the electronic rearview mirror screen is consistent with the human eye's visual range; The implementation method of the intelligent electronic rearview mirror system comprises the following steps: Step 1: Build a human eye attention detection network. Place a monocular camera in front of the driver to take a photo of the driver's face and input it into the network to obtain the driver's head posture and line of sight direction. Step 2: When the driver looks to the left, the display screen installed at the left A-pillar position displays the left side and rear images of the car captured by the camera on the left rearview mirror and adjusts the display angle from angle a to angle a1 to increase the display range of the left side and rear images of the car. The central control LCD display screen displays the left lane scene. After the driver's sight returns to the right, the central control LCD keeps displaying. Step 3, when the driver's line of sight is to the right, the display screen installed at the right A-pillar position displays the right side and rear side images of the car captured by the camera on the right rearview mirror and adjusts the display angle from angle b to angle b1 to increase the display range of the right side and rear side images of the car, and the central control LCD display screen displays the right lane scene. After the line of sight returns to the right side, the central control LCD display screen keeps displaying; Step 4: Build a target detection network to detect targets in the vehicle's surrounding environment. When the driver performs operations such as changing lanes or turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays dangerous objects around the vehicle.
2. The intelligent electronic rearview mirror system based on human eye attention according to claim 1, characterized in that: The target detection unit uses a target detection network to detect and classify surrounding objects. The target detection network specifically includes the following sub-steps: A target detection network based on YOLOv3 is constructed. By combining the KITTI public dataset and the self-sampling dataset, the vehicle runtime environment targets are divided into five categories: person, bike, car, truck, and bus. The target detection network is trained.
3. The intelligent electronic rearview mirror system based on human eye attention according to claim 1, characterized in that: The display unit displays the image captured by the camera on the electronic rearview mirror screen and the central control LCD; at the same time, it detects targets in the vehicle's surrounding environment. When the driver performs operations such as changing lanes and turning, or when dangerous targets approach, the left and right electronic rearview mirror screens display the detected objects in the rear area, and the central control LCD displays the dangerous objects around the vehicle.
4. The intelligent electronic rearview mirror system based on human eye attention according to claim 1, characterized in that: The human eye attention detection unit uses MTCNN to perform accurate face detection, obtains the driver's head posture through the head posture network, obtains the driver's line of sight direction using the line of sight detection network, and finally obtains the driver's attention direction by combining the head posture and the line of sight direction; specifically includes: The human eye attention detection network takes the driver image captured by the monocular camera as input and outputs the driver's attention direction; The human eye attention detection network first uses the MTCNN multi-task convolutional neural network to extract the face of the driver image captured by the monocular camera to obtain the face area; Then, the face area is detected using the head posture estimation algorithm to obtain the Euler angles (yaw angle, pitch angle, roll angle) of the head posture. At the same time, the face area is input into the sight line detection neural network to obtain the driver's sight line direction. Finally, the driver's line of sight direction is converted into Euler angles and combined with the Euler angles of the head posture for calibration, and finally the driver's attention direction is obtained.
5. The intelligent electronic rearview mirror system based on human eye attention according to claim 4, characterized in that: The adjustment unit uses the driver's head posture and line of sight direction obtained by the human eye attention detection unit to continuously adjust the display content of the electronic rearview mirror on the corresponding side according to the driver's attention direction; at the same time, according to the driver's attention direction, the lane information on different sides is displayed on the central control LCD display screen.
6. The intelligent electronic rearview mirror system based on human eye attention according to claim 1, characterized in that: The human eye attention detection network in step 1 takes the driver image captured by the monocular camera as input and outputs the driver's attention direction; The human eye attention detection network first uses the MTCNN multi-task convolutional neural network to extract the face of the driver image captured by the monocular camera to obtain the face area; Then, the face area is detected using the head posture estimation algorithm to obtain the Euler angles of the head posture, namely the yaw angle, pitch angle, and roll angle. At the same time, the face area is input into the sight line detection neural network to obtain the driver's sight line direction. Finally, the driver's line of sight direction is converted into Euler angles and combined with the Euler angles of the head posture for calibration, and finally the driver's attention direction is obtained.
