A method and system for detecting fatigue driving under low illumination

Abstract

The present invention relates to a method and system for detecting fatigue driving under low illumination, comprising the following steps: obtaining a vehicle driving speed, and when the vehicle driving speed is 0, continuously collecting a driver's face image, and extracting a human eye spot in the driver's face image; calculating the maximum average brightness of the human eye spot, and extracting the original contour of the human eye spot under the maximum average brightness; longitudinally dividing the original contour into a first light spot area, a second light spot area, and a third light spot area, and calculating the average brightness of the first light spot area, the second light spot area, and the third light spot area of ​​the human eye spot under the maximum average brightness; calculating the current brightness of the first light spot area, the second light spot area, and the third light spot area, and determining whether the driver is driving with severe fatigue.

Description

Technical Field

[0001] The present invention relates to the field of fatigue driving detection methods, and in particular to a method and system for detecting fatigue driving under low illumination. Background Art

[0002] Fatigue driving refers to the phenomenon that the driver's physiological and psychological functions are disturbed after driving for a long time, and the driving skills are objectively reduced. Drivers with poor or insufficient sleep are prone to fatigue after driving for a long time.

[0003] In the prior art, the driver's eyes are generally analyzed through image detection technology, and the width-to-height ratio of the eyes is analyzed to determine whether the driver is squinting; or the ratio of the blinking time to the opening time of the human eye is analyzed to determine whether the driver is dozing off or closing his eyes for a long time; or the driver's head is turned to determine whether the driver can look left and right sensitively to determine the road conditions, and finally determine whether the driver is driving fatigued. In this solution, the analysis is based on the collection of facial images. In low-light conditions, in order to obtain a sufficiently clear image, complex image enhancement, image brightening and other processing are required, or infrared light is emitted to assist in extracting a clear image.

[0004] The purpose of the present invention is to design a method and system for detecting fatigue driving under low illumination in view of the above-mentioned problems in the prior art. Summary of the invention

[0005] In view of the problems existing in the above-mentioned prior art, the present invention provides a method and system for detecting fatigue driving under low illumination, which can effectively solve at least one problem existing in the above-mentioned prior art.

[0006] The technical solution of the present invention is:

[0007] A method for detecting fatigue driving under low illumination comprises the following steps:

[0008] Obtaining the vehicle's driving speed, and when the vehicle's driving speed is 0, continuously collecting the driver's face image, and extracting the human eye spot in the driver's face image;

[0009] Calculating the maximum average brightness of the human eye light spot, and extracting the original outline of the human eye light spot at the maximum average brightness;

[0010] The original outline is longitudinally divided into a first light spot area, a second light spot area, and a third light spot area, and the average brightness of the first light spot area, the second light spot area, and the third light spot area of ​​the human eye light spot at the maximum average brightness is calculated;

[0011] Calculate the current brightness of the first light spot area, the second light spot area, and the third light spot area,

[0012] If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued;

[0013] If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue;

[0014] If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue;

[0015] If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

[0016] Further, extracting the eye spots in the driver's face image includes:

[0017] Collect the image in front of the vehicle and generate a spot image through the human eye spot neural network;

[0018] Extracting a contour from the light spot image to obtain a light spot contour;

[0019] The eye light spot is obtained by performing contour matching on the light spot contour in the face image.

[0020] Furthermore, the human eye spot neural network is obtained by the following steps:

[0021] The image in front of the vehicle is taken as input, and the human eye spot is taken as output, and a neural network is trained to obtain a human eye spot neural network.

[0022] Furthermore, the first time is 1.5s.

[0023] Further, before calculating the current brightness of the first light spot area, the second light spot area, and the third light spot area, executing:

[0024] The shape of the original contour in the third light spot area is extracted, shape matching is performed in the human eye light spot through the shape of the original contour in the third light spot area to obtain the bottom position of the original contour, and the original contour is placed in the face image at the bottom position.

[0025] Further, after extracting the eye spot in the driver's face image and before calculating the average brightness of the eye spot, the following is performed:

[0026] The human eye light spot is binarized with a preset grayscale threshold to obtain a binarized human eye light spot image, and the binarized human eye light spot image is used as a mask to expose the human eye light spot.

[0027] Furthermore, after extracting the original outline of the human eye spot, execute:

[0028] The original contour is smoothed.

[0029] Furthermore, the driver's face image is continuously collected in front and above the driver's face.

[0030] A low-light fatigue driving detection system is further provided, comprising the following modules:

[0031] The eye spot extraction module is used to obtain the vehicle speed. When the vehicle speed is 0, the driver's face image is continuously collected to extract the eye spot in the driver's face image.

[0032] A brightness calculation module, used to calculate the maximum average brightness of the human eye light spot and extract the original outline of the human eye light spot at the maximum average brightness;

[0033] A reference spot signal calculation module is used to divide the original profile into a first spot area, a second spot area, and a third spot area on average longitudinally, and calculate the average brightness of the first spot area, the second spot area, and the third spot area of ​​the human eye spot at the maximum average brightness;

[0034] A fatigue driving judgment module is used to calculate the current brightness of the first light spot area, the second light spot area, and the third light spot area.

[0035] If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued;

[0036] If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue;

[0037] If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue;

[0038] If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

[0039] Therefore, the present invention provides the following effects and / or advantages:

[0040] This application extracts and analyzes the eye spot, quantifies the effect of eye fatigue and different degrees of closure on the eye spot, and thus determines the driver's fatigue. This application can detect the driver's fatigue in low light, without extracting the image of the face and performing complex brightening operations. The eye spot itself has higher brightness than other areas of the face and has a certain shape, and is extracted to determine fatigue driving.

[0041] The present application generates a spot image through a human eye spot neural network, and performs image calculation on the spot image to match the human eye spot in the face image, so as to accurately find the human eye spot and eliminate interference such as glasses reflection.

[0042] The present application uses the shape of the original outline in the third light spot area to find the position of the bottom of the human eye light spot in the image, which is used to locate the position where the original outline matches the human eye light spot in different situations.

[0043] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and the drawings.

[0044] It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are intended to provide further explanation of the invention as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic diagram of a flow chart provided for an embodiment of the present invention.

[0046] Figure 2 Schematic diagram of a normally open human eye and its light spot.

[0047] Figure 3 for Figure 2 Schematic diagram of the original outline of the human eye light spot is shown.

[0048] Figure 4 Schematic diagram of the first light spot area, the second light spot area, and the third light spot area.

[0049] Figure 5 Schematic diagram of a slightly closed human eye and its light spot.

[0050] Figure 6 Schematic diagram of the moderate closure of the human eye and its light spot.

[0051] Figure 7 It is a schematic diagram of identifying the position of the human eye light spot in the human eye through the original contour image of the third light spot area. DETAILED DESCRIPTION

[0052] In order to facilitate understanding by those skilled in the art, the present invention is now described in further detail by way of examples:

[0053] refer to Figure 1 , a method for detecting fatigue driving under low illumination, comprising the following steps:

[0054] S1, obtaining the vehicle speed, and when the vehicle speed is 0, continuously collecting the driver's face image, and extracting the eye spot in the driver's face image;

[0055] This embodiment only starts fatigue driving detection when the vehicle speed is 0. This is because low illumination generally occurs when the vehicle is driving on mountain roads, highways and other road conditions. Under such road conditions, less light enters the interior of the vehicle from the outside. The driver's face image captured by the camera inside the vehicle is difficult to distinguish the shape of the eyes in the face, and it is impossible to use existing technical means such as analyzing eye width and blinking time for analysis. This embodiment uses the fact that the headlights of the vehicle illuminate the road obliquely downward in front of the vehicle at night, generating a wide illumination area. After the light enters the human eye, it will generate a human eye spot on the surface of the human eye under the reflection of the human eye. At this time, the vehicle is stationary, and the light reflected back into the vehicle by the object in front of the vehicle is basically unchanged. The human eye spot is related to the opening degree of the human eye. Therefore, capturing the human eye spot can be used for fatigue driving judgment in subsequent steps when the human eye contour cannot be distinguished under low illumination, etc., through the human eye spot with brightness significantly greater than the surrounding.

[0056] S2, calculating the maximum average brightness of the human eye light spot, and extracting the original contour of the human eye light spot at the maximum average brightness;

[0057] In this step, when the driver is tired, the driver is not asleep, and is in a cycle of opening eyes normally and gradually closing eyes, or when the driver is normal, his eyes are normally open. Therefore, the maximum average brightness of the human eye spot within a certain time range is captured, so that the human eye spot when the driver looks forward normally can be obtained. Figure 2 The eyes are normally open, and then Figure 2 The human eye spot is obtained from the edge calculation, and finally the human eye spot is obtained as follows Figure 3 The original outline of the human eye spot is shown.

[0058] Then, edge detection is performed based on the average brightness of the eye spot of the driver's face image, so as to obtain the original outline of the eye spot. The edge detection can be implemented by the existing Canny operator.

[0059] S3, dividing the original outline into a first light spot area, a second light spot area, and a third light spot area on average in the longitudinal direction, and calculating the average brightness of the first light spot area, the second light spot area, and the third light spot area of ​​the human eye light spot at the maximum average brightness;

[0060] In this embodiment, the area is divided into three areas from top to bottom. Figure 4 As shown, from top to bottom, they are the first light spot area, the second light spot area, and the third light spot area. Figure 4 The area shown is Figure 3 Calculate the average brightness of the three spot areas and get Figure 2The average brightness of the first light spot area of ​​the human eye light spot shown is 145.27, the average brightness of the second light spot area is 162.81, and the average brightness of the third light spot area is 77.12.

[0061] S4, calculating current brightness of the first light spot area, the second light spot area, and the third light spot area.

[0062] At this time, the average brightness of the three areas at the maximum brightness is obtained in the above, and the three average brightnesses are used as the benchmarks of the three areas. When the human body is tired, the degree of opening of the human eye decreases, and the degree of human fatigue is positively correlated with the degree of opening of the human eye. It can be understood that the more tired the human body is, the harder it is to open the human eye. Since the present application is suitable for execution under low illumination, the light spot formed by the reflection of the wider lighting area in front of the vehicle in the human eye is located below the human eye. In the process of the gradual closing of the human eye, the eyelashes of the human eye are gradually blocked from the first light spot area to the third light spot area, and the eyelids of the human eye are also gradually blocked from the first light spot area to the third light spot area, and the brightness of the human eye light spot is positively correlated with the amount of opening of the human eye. Therefore, the brightness changes of the first light spot area, the second light spot area, and the third light spot area of ​​the human eye light spot can be used to judge the amount of opening of the human eye through the following judgment method.

[0063] If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued;

[0064] When the human eyes are normally open, the first light spot area, the second light spot area, and the third light spot area should all reach their maximum. Therefore, under this judgment method, it can be judged that the driver's eyes are normally open and he is not driving fatigued.

[0065] If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue;

[0066] When the eyes are slightly closed, Figure 5 As shown, eyelashes will block the first light spot area, so that its current brightness is greatly reduced. Then, as the degree of opening of the human eye decreases, the amount of external light entering the human eye decreases accordingly, so that the brightness of the entire human eye light spot is reduced as a whole. Figure 5As shown, the average brightness of the first light spot area of ​​the human eye light spot is 125.37, the average brightness of the second light spot area is 135.36, and the average brightness of the third light spot area is 64.28, which is consistent with the brightness variation range of mild fatigue driving. Therefore, it is determined that the driver is driving with mild fatigue.

[0067] If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue;

[0068] Similarly, when the human eye is further closed, Figure 6 As shown, eyelashes and eyelids will further block the first light spot area, and at the same time eyelashes block the second light spot area, and the brightness of the entire eye light spot is further reduced, so that the driver can be judged to be driving with moderate fatigue.

[0069] If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

[0070] Similarly, when the human eyes are further closed, the light spot of the human eyes has been largely blocked and has lost its original shape and brightness. At this time, the above quantification can be used to determine whether the driver is driving with severe fatigue.

[0071] Further, extracting the eye spots in the driver's face image includes:

[0072] Collect the image in front of the vehicle and generate a spot image through the human eye spot neural network;

[0073] Extracting a contour from the light spot image to obtain a light spot contour;

[0074] The eye light spot is obtained by performing contour matching on the light spot contour in the face image.

[0075] In this step, since the human body may wear glasses, etc., reflections may occur, and it is difficult to distinguish the position of the human eyes in low light. The reflections of other objects are easily misjudged as human eye spots. Therefore, this step is used to identify the human eye spots.

[0076] Furthermore, the human eye spot neural network is obtained by the following steps:

[0077] The image in front of the vehicle is taken as input, and the human eye spot is taken as output, and a neural network is trained to obtain a human eye spot neural network.

[0078] In this step, the neural network can be trained by this training method, so that the neural network finds the relationship between the image in front of the vehicle and the human eye spot, such as distortion, color, brightness, projection, etc., and outputs the calculated human eye spot image for the above-mentioned contour matching, etc., to accurately identify the position of the human eye spot and reduce external interference. Specifically, the human eye spot in the face photo can be manually circled as the output of the neural network, so as to train a neural network that can convert the image into the human eye spot.

[0079] Furthermore, the first time is 1.5s.

[0080] Since the human body normally has slight physiological movements, by setting the duration to 1.5s, the interference of normal human movements such as rubbing eyes and blinking on judgment can be reduced.

[0081] Further, before calculating the current brightness of the first light spot area, the second light spot area, and the third light spot area, executing:

[0082] The shape of the original contour in the third light spot area is extracted, shape matching is performed in the human eye light spot through the shape of the original contour in the third light spot area to obtain the bottom position of the original contour, and the original contour is placed in the face image at the bottom position.

[0083] When the human eye is closed to different degrees, the first light spot area and the second light spot area are easily partially or completely blocked, so that the algorithm cannot identify the position of the human eye light spot, resulting in calculation errors. Since the third light spot area is located at the bottom of the human eye, the influence of different degrees of closure when the human eye is tired on the third light spot area is the smallest. This step extracts the contour of the shape in the third light spot area, so as to Figure 5 Find the position where the third spot area matches in the human eye spot shown in the figure, Figure 7 Then, the original outline is placed into the human eye light spot based on the position of the third light spot area, so as to obtain the correct position of the light spot.

[0084] Further, after extracting the eye spot in the driver's face image and before calculating the average brightness of the eye spot, the following is performed:

[0085] The human eye light spot is binarized with a preset grayscale threshold to obtain a binarized human eye light spot image, and the binarized human eye light spot image is used as a mask to expose the human eye light spot.

[0086] Since the human eye may be affected by weaker light such as that in front of the vehicle, it may produce extra eye spots in the human eye, affecting subsequent judgment. Therefore, in this step, the image can be binarized first to eliminate the influence of weaker light and only retain the image of the area with correct eye spots.

[0087] Furthermore, after extracting the original outline of the human eye spot, execute:

[0088] The original contour is smoothed.

[0089] In this embodiment, the original contour image is as follows: Figure 3 As shown, the edge burrs are numerous and complex, so the original contour image can be smoothed by smoothing, making it easier to proceed with subsequent steps.

[0090] Furthermore, the driver's face image is continuously collected in front and above the driver's face.

[0091] A low-light fatigue driving detection system is further provided, comprising the following modules:

[0092] The eye spot extraction module is used to obtain the vehicle speed. When the vehicle speed is 0, the driver's face image is continuously collected to extract the eye spot in the driver's face image.

[0093] A brightness calculation module, used to calculate the maximum average brightness of the human eye light spot and extract the original outline of the human eye light spot at the maximum average brightness;

[0094] A reference spot signal calculation module is used to divide the original profile into a first spot area, a second spot area, and a third spot area on average longitudinally, and calculate the average brightness of the first spot area, the second spot area, and the third spot area of ​​the human eye spot at the maximum average brightness;

[0095] A fatigue driving judgment module is used to calculate the current brightness of the first light spot area, the second light spot area, and the third light spot area.

[0096] If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued;

[0097] If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue;

[0098] If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue;

[0099] If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

[0100] The working principle of the fatigue driving detection system of this embodiment is the same as that of the above-mentioned fatigue driving detection method.

[0101] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0103] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0104] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0105] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms should not be understood as necessarily being directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

Claims

1. A method for detecting fatigue driving under low illumination, characterized in that: The following steps are involved: Obtaining the vehicle's driving speed, and when the vehicle's driving speed is 0, continuously collecting the driver's face image, and extracting the human eye spot in the driver's face image; Calculating the maximum average brightness of the human eye light spot, and extracting the original outline of the human eye light spot at the maximum average brightness; The original outline is longitudinally divided into a first light spot area, a second light spot area, and a third light spot area, and the average brightness of the first light spot area, the second light spot area, and the third light spot area of ​​the human eye light spot at the maximum average brightness is calculated; Calculate the current brightness of the first light spot area, the second light spot area, and the third light spot area, If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued; If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue; If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue; If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

2. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: Extracting the eye spots in the driver's face image includes: Collect the image in front of the vehicle and generate a spot image through the human eye spot neural network; Extracting a contour from the light spot image to obtain a light spot contour; The eye light spot is obtained by performing contour matching on the light spot contour in the face image.

3. A method for detecting fatigue driving under low illumination according to claim 2, characterized in that: The human eye spot neural network is obtained through the following steps: The image in front of the vehicle is taken as input, and the human eye spot is taken as output, and a neural network is trained to obtain a human eye spot neural network.

4. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: The first time is 1.5s.

5. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: Before calculating the current brightness of the first light spot area, the second light spot area, and the third light spot area, execute: The shape of the original contour in the third light spot area is extracted, shape matching is performed in the human eye light spot through the shape of the original contour in the third light spot area to obtain the bottom position of the original contour, and the original contour is placed in the face image at the bottom position.

6. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: After extracting the eye spots in the driver's face image and before calculating the average brightness of the eye spots, execute: The human eye light spot is binarized with a preset grayscale threshold to obtain a binarized human eye light spot image, and the binarized human eye light spot image is used as a mask to expose the human eye light spot.

7. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: After extracting the original outline of the human eye spot, execute: The original contour is smoothed.

8. The method for detecting fatigue driving under low illumination according to claim 1, characterized in that: The driver's face image is continuously collected in front and above the driver's face.

9. A low-light fatigue driving detection system, characterized in that: Includes the following modules: The eye spot extraction module is used to obtain the vehicle speed. When the vehicle speed is 0, the driver's face image is continuously collected to extract the eye spot in the driver's face image. A brightness calculation module, used to calculate the maximum average brightness of the human eye light spot and extract the original outline of the human eye light spot at the maximum average brightness; A reference spot signal calculation module is used to divide the original profile into a first spot area, a second spot area, and a third spot area on average longitudinally, and calculate the average brightness of the first spot area, the second spot area, and the third spot area of ​​the human eye spot at the maximum average brightness; A fatigue driving judgment module is used to calculate the current brightness of the first light spot area, the second light spot area, and the third light spot area. If the current brightness of the first light spot area is greater than 90% of the average brightness of the first light spot area, the current brightness of the second light spot area is greater than 90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is greater than 90% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is not driving fatigued; If the current brightness of the first light spot area is 60%-80% of the average brightness of the first light spot area, the current brightness of the second light spot area is 80%-90% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 80%-90% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with mild fatigue; If the current brightness of the first light spot area is 40%-60% of the average brightness of the first light spot area, the current brightness of the second light spot area is 60%-80% of the average brightness of the second light spot area, and the current brightness of the third light spot area is 70%-80% of the average brightness of the third light spot area, and it lasts for more than the first time, it is determined that the driver is driving with moderate fatigue; If the current brightness of the first light spot area is less than 40% of the average brightness of the first light spot area, the current brightness of the second light spot area is less than 60% of the average brightness of the second light spot area, and the current brightness of the third light spot area is less than 70% of the average brightness of the third light spot area, and it lasts for more than a first time, it is determined that the driver is driving with severe fatigue.

Images