Method for detecting lane departure or illegal lane change in tunnel based on eye features

Abstract

The invention discloses a method for detecting lane departure or illegal lane change in a tunnel based on eye features, and the method comprises the following steps: S1, tracking the position of a pupil region of a driver in real time through an infrared pupil tracker, S2, obtaining the pupil area of the driver through a pupil image captured by an infrared camera, and S3, determining the pupil area of the driver according to the pupil area. S3, judging whether the vehicle enters the tunnel or not according to the change of the pupil area of the driver; and S4, judging whether the vehicle has lane departure or illegal lane change behaviors or not according to the change of the pupil area of the driver. According to the method, the pupil position of the driver is captured through the infrared pupil tracker, then the pupil area of the driver is obtained, the pupil change rule is obtained according to the change of the pupil area when the vehicle enters the tunnel or does not enter the tunnel, and then whether the vehicle has lane departure or illegal lane change behaviors or not can be detected very accurately.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic safety, and in particular relates to a method for detecting lane deviation or illegal lane change in a tunnel based on eye features. Background Art

[0002] Tunnels are common infrastructure in the transportation field. Traffic accidents caused by lane departure and illegal lane changes are numerous. Due to limited vision and light effects in tunnels, drivers may deviate from lanes or change lanes illegally, which increases the risk of traffic accidents. Usually, we use lane line detection to determine whether a vehicle has deviated or changed lanes illegally in a tunnel.

[0003] However, due to factors such as insufficient light and humid air in tunnels, the high-definition camera used for lane line detection cannot capture lane line images well in dim light or worn lane lines. Therefore, lane line detection technology is difficult to be well applied to detect lane departure and illegal lane change in some special tunnel environments, which will lead to inaccurate detection results.

[0004] In view of this, the present invention is proposed. Summary of the invention

[0005] In order to solve the problems in the prior art, the present invention proposes a method for detecting lane deviation or illegal lane change in a tunnel based on eye features, the purpose of which is to use the lighting facilities arranged in the tunnel to reversely infer whether the driver has lane deviation or illegal lane change by whether there is an abnormality in the change pattern of the driver's pupil area. When a vehicle is driving in a tunnel and the lane deviates or changes lanes illegally, the light intensity to the driver's eyes will change during normal trajectory driving, thereby causing changes in the size of the pupil and the area of ​​the pupil to be opened. By collecting data on the size of the pupil and the area of ​​the pupil to be opened, it is determined whether a lane deviation or illegal lane change has occurred.

[0006] A method for detecting lane deviation or illegal lane change in a tunnel based on eye features of the present invention comprises the following steps:

[0007] S1. Use an infrared pupil tracker to track the driver's pupil area in real time.

[0008] S2, obtain the driver's pupil area by using the pupil image captured by the infrared camera,

[0009] S3, judging whether the vehicle has entered a tunnel based on the change in the driver's pupil area,

[0010] S4. Determine whether the vehicle has lane deviation or illegal lane changing behavior based on changes in the driver's pupil area.

[0011] Furthermore, in step S1, before using the infrared pupil tracker to measure the driver's pupil area in real time, the infrared pupil tracker needs to be calibrated, and then the driver's pupil area image is captured in real time by the infrared camera.

[0012] Furthermore, in step S2, the pupil area of ​​the driver is obtained by using the pupil image captured by the infrared camera, which specifically includes:

[0013] The pupil image captured by the infrared camera uses an edge detection algorithm or a threshold-based segmentation algorithm to identify and depict the outline of the pupil, and the driver's pupil area is obtained by calculating the area within the outline.

[0014] Further, in step S3, judging whether the vehicle enters a tunnel according to the change in the pupil area of ​​the driver specifically includes:

[0015] Assume that A(t) represents the pupil area at time t. Before entering the tunnel, the pupil area will maintain a stable value A0. During the process of entering the tunnel, the pupil area changes. The maximum value of the pupil area is set to A. max , the minimum value is A min ,

[0016] When the variation pattern of A(t) increases first and then decreases, and A max >2A min When the vehicle enters the tunnel,

[0017] Otherwise, the vehicle has not entered the tunnel.

[0018] Furthermore, in step S4, determining whether the vehicle has lane departure or illegal lane change behavior includes:

[0019] When a vehicle deviates from a lane or changes lanes illegally, the intensity of light received by the driver's eyes will change, causing the pupil area to change accordingly. The change in pupil area can be used to determine whether the vehicle has deviated from a lane or changed lanes illegally.

[0020] Assuming that A(t) represents the pupil area at time t, and Q(t) represents the light intensity in the tunnel at time t, the functional relationship between A(t) and Q(t) is expressed as:

[0021] A(t)=F(Q(t)),

[0022] Among them, A(t) is negatively correlated with Q(t), and A(t) decreases as Q(t) increases.

[0023] Furthermore, the size of Q(t) in the tunnel is related to the position of the vehicle in the tunnel. The closer to the center of the road, the larger Q(t) is, and the closer to the sides of the tunnel, the smaller Q(t) is.

[0024] Assuming that the distance from the vehicle to the left side of the tunnel at time t is d1, and the distance from the vehicle to the right side of the tunnel is d2, the relationship between Q(t) and d1 and d2 can be expressed as:

[0025] Q(t)=Z(|d1-d2|)

[0026] Q(t) is negatively correlated with the size of |d1-d2|. When |d1-d2| is 0, it means that the vehicle is driving in the center of the tunnel. When |d1-d2| is larger, it means that the vehicle is closer to the sides of the tunnel.

[0027] A(t) is positively correlated with the size of |d1-d2|. When |d1-d2| is 0, A(t) is the minimum value. The larger |d1-d2| is, the larger A(t) is.

[0028] Furthermore, the changes in the driver's pupil area and the presence of lane departure or illegal lane change are as follows:

[0029] When A(t) fluctuates within a certain range, and the maximum value of the fluctuation range is A max >1.2A0, minimum value A min When the value is less than 0.8A0, the vehicle is judged to have lane deviation.

[0030] When A(t) increases first and then decreases, and its final value tends to the stable value A0, and this change process lasts for more than 3 seconds, it is judged that the vehicle is changing lanes illegally.

[0031] Compared with the prior art, the present invention provides a method for detecting lane deviation or illegal lane change in a tunnel based on eye features. By using an infrared pupil tracker to capture the driver's pupil position, the driver's pupil area is obtained. The pupil change law is obtained based on the change in pupil area when the vehicle enters the tunnel and when it does not enter the tunnel, and thus it is possible to very accurately detect whether the vehicle has lane deviation or illegal lane change behavior. In addition, the design and operation of the infrared pupil tracker used in the present invention ensure that the pupil area can be accurately and continuously measured even in low light or no light conditions, and real-time pupil area data can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of a method for detecting lane deviation or illegal lane change in a tunnel based on eye features of the present invention. DETAILED DESCRIPTION

[0033] The present invention is further explained below in conjunction with the drawings and specific embodiments of the specification. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0034] According to an embodiment of the present invention, a method for detecting lane departure or illegal lane change based on eye features under tunnel light distribution is provided, the purpose of which is to utilize the lighting facilities arranged in the tunnel to reversely infer whether the driver has lane departure or illegal lane change by determining whether there is any abnormality in the change pattern of the driver's pupil area.

[0035] Specifically, in experimental studies, the size of pupil area changes under different lighting conditions. The study found that as the illuminance increases, the eye opening area and pupil size change from large to small. When the driver is driving in a tunnel, the effect of tunnel light on the driver's pupil area follows a certain law of change. When the driver enters a darker tunnel from a bright external environment, in order to adapt to the lower light level, the pupil will naturally expand, that is, the pupil area will increase, so that more light can enter the eye and improve vision in a dark environment; as the driver drives in the tunnel, the pupil will gradually adjust to the state of adapting to the current light level. Generally speaking, when driving in a tunnel, the light is most abundant when driving in the center of the tunnel, and the light intensity will gradually decrease from the center to the sides. Therefore, the closer the vehicle is to the center, the smaller the driver's pupil area is, and the closer it is to the sides, the larger the pupil area is. When the driver is about to leave the tunnel and enter a bright external environment, the pupil area will decrease rapidly to reduce the amount of light entering the eye and avoid glare and discomfort caused by strong light.

[0036] The above-mentioned change in pupil area is controlled by the autonomic nervous system and is an unconscious physiological reaction. Based on this physiological reaction, the change pattern of the driver's pupil caused by the change of tunnel light can be used to reversely infer whether the driver has deviated from the lane or changed lanes illegally.

[0037] Example 1

[0038] According to the embodiment of the present invention, in combination with the appended claims of the present invention, Figure 1 The present invention provides a method for detecting lane departure or illegal lane change in a tunnel based on eye features, comprising the following steps:

[0039] Step 1: Use an infrared pupil tracker to track the position of the driver's pupil area in real time.

[0040] An infrared pupil tracker is used to track the position of the driver's pupil in the tunnel, that is, an infrared camera is used to capture high-resolution images of the driver's eyes, which pay special attention to the pupil area. It should be noted that before using the infrared pupil tracker to measure the driver's pupil area in real time, the infrared pupil tracker needs to be calibrated, and then the driver's pupil area image is captured in real time by the infrared camera. The design and operation of the infrared pupil tracker ensures that the pupil area can be accurately and continuously measured even in low light or no light conditions to obtain real-time pupil area data.

[0041] Step 2: Obtain the driver's pupil area by using the pupil image captured by the infrared camera.

[0042] The images captured by the infrared pupil tracker are analyzed by specific image processing algorithms that are able to identify the edge of the pupil, using edge detection algorithms or threshold-based segmentation techniques to identify the outline of the pupil. Once the edge of the pupil is detected, the algorithm will outline the pupil. After the outline of the pupil is identified, the algorithm will calculate the area of ​​the area within the outline to obtain the driver's pupil area.

[0043] Step 3: Determine whether the vehicle has entered a tunnel based on changes in the driver's pupil area.

[0044] When entering or exiting a tunnel, the difference in brightness inside and outside the tunnel may cause the driver to experience a "black hole" or "white hole" effect. At this time, the pupil area will change drastically compared to lane deviation or illegal lane change. After the drastic change, it will return to a stable value. At this time, it is judged that the vehicle has entered or left the tunnel. When the driver passes through and enters the tunnel, due to the black hole phenomenon, the real-time pupil area changes drastically, and then tends to a stable value. At this time, it is judged that the vehicle has entered the tunnel and is driving normally.

[0045] Assume that A(t) represents the pupil area at time t. Before entering the tunnel, the pupil area will maintain a stable value A0. During the process of entering the tunnel, the pupil area changes dramatically. The maximum value of the pupil area is set to A max , the minimum value is A min When A(t) changes first and then decreases, and A max >2A min , it is judged as entering a tunnel.

[0046] Step 4: Determine whether the vehicle has lane deviation or illegal lane changing behavior based on changes in the driver's pupil area.

[0047] According to the influence of the light changes on the pupil size in the tunnel, the real-time pupil area change can be used to infer whether the vehicle has lane deviation or illegal lane change in the tunnel, and the change of the pupil area during lane deviation and illegal lane change can be used to distinguish whether the vehicle has lane deviation or illegal lane change. The details are as follows:

[0048] After the pupil area changes due to the black hole phenomenon when entering the tunnel, the pupil area will remain at a stable value A0 during normal driving. When the vehicle deviates from the lane or changes lanes illegally, the intensity of light received by the driver's eyes will change, causing the pupil area to change accordingly. Based on the change in pupil area, it can be inferred whether the vehicle has deviated from the lane or changed lanes illegally.

[0049] A(t) represents the pupil area at time t, Q(t) represents the light intensity in the tunnel at time t, and the relationship between A(t) and Q(t) can be expressed as:

[0050] A(t)=f(Q(t))

[0051] A(t) is negatively correlated with Q(t), and A(t) decreases as Q(t) increases.

[0052] The size of Q(t) in the tunnel is related to the position of the vehicle in the tunnel. The closer to the center of the road, the larger Q(t), and the closer to the sides of the tunnel, the smaller Q(t). Assuming that the distance from the left side of the vehicle at time t is d1, and the distance from the right side is d2, the relationship between Q(t) and d1 and d2 can be expressed as:

[0053] Q(t)=Z(|d1-d2|)

[0054] Q(t) is negatively correlated with the size of |d1-d2|. When |d1-d2| is 0, the vehicle is judged to be driving in the center of the tunnel. The larger |d1-d2| is, the closer the vehicle is to the sides of the tunnel.

[0055] It is found that the relationship between A(t) and |d1-d2| is positively correlated. When |d1-d2| is 0, A(t) is the minimum value. The larger |d1-d2| is, the larger A(t) is.

[0056] From this, we can infer that when A(t) changes, the distance between the vehicle and the two sides of the tunnel will change, that is, |d1-d2| will change, so as to judge whether the vehicle has lane deviation or illegal lane change. Lane change means that the entire lateral displacement must exceed 3 meters. If the lane deviates, it is within the lane, and the lateral displacement generally does not exceed 1.5 meters, showing a lateral back-and-forth swing. According to their respective characteristics, different pupil change rules are used to distinguish them. The specific judgment is based on the following rules:

[0057] When A(t) fluctuates within a certain range, and the maximum value of the fluctuation range is A max >1.2A0, minimum value A min <0.8A0, it is judged that the vehicle has deviated from the lane.

[0058] When A(t) increases first and then decreases and finally tends to a stable value A0, and this process lasts for more than 3 seconds, it is judged that the vehicle is changing lanes illegally.

[0059] According to an embodiment of the present invention, when a vehicle deviates from a lane or changes lanes illegally in a tunnel, the intensity of light received by the driver will change, thereby affecting the change in pupil area. According to the above specific implementation scheme and the rules, it can be determined whether the vehicle deviates from a lane or changes lanes illegally in a tunnel.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting lane departure or illegal lane change in a tunnel based on eye features, characterized in that: The following steps are involved: S1. Use an infrared pupil tracker to track the driver's pupil area in real time. S2, obtain the driver's pupil area by using the pupil image captured by the infrared camera, S3, judging whether the vehicle has entered a tunnel based on the change in the driver's pupil area, S4. Determine whether the vehicle has lane deviation or illegal lane changing behavior based on changes in the driver's pupil area.

2. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 1, characterized in that: In step S1, before using the infrared pupil tracker to measure the driver's pupil area in real time, the infrared pupil tracker needs to be calibrated, and then the driver's pupil area image is captured in real time by the infrared camera.

3. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 1, characterized in that: In step S2, the pupil area of ​​the driver is obtained by using the pupil image captured by the infrared camera, which specifically includes: The pupil image captured by the infrared camera uses an edge detection algorithm or a threshold-based segmentation algorithm to identify and depict the outline of the pupil, and the driver's pupil area is obtained by calculating the area within the outline.

4. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 1, characterized in that: In step S3, judging whether the vehicle has entered a tunnel according to the change in the pupil area of ​​the driver specifically includes: Assume that A(t) represents the pupil area at time t. Before entering the tunnel, the pupil area will maintain a stable value A0. During the process of entering the tunnel, the pupil area changes. The maximum value of the pupil area is set to A. max , the minimum value is A min , When the variation pattern of A(t) increases first and then decreases, and A max >2A min When the vehicle enters the tunnel, Otherwise, the vehicle has not entered the tunnel.

5. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 1, characterized in that: In step S4, determining whether the vehicle has lane departure or illegal lane change behavior includes: When a vehicle deviates from a lane or changes lanes illegally, the intensity of light received by the driver's eyes will change, causing the pupil area to change accordingly. The change in pupil area can be used to determine whether the vehicle has deviated from a lane or changed lanes illegally. Assuming that A(t) represents the pupil area at time t, and Q(t) represents the light intensity in the tunnel at time t, the functional relationship between A(t) and Q(t) is expressed as: A(t)=F(Q(t)), Among them, A(t) is negatively correlated with Q(t), and A(t) decreases as Q(t) increases.

6. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 5, characterized in that: The size of Q(t) in the tunnel is related to the position of the vehicle in the tunnel. The closer to the center of the road, the larger Q(t) is, and the closer to the sides of the tunnel, the smaller Q(t) is. Assuming that the distance from the vehicle to the left side of the tunnel at time t is d1, and the distance from the vehicle to the right side of the tunnel is d2, the relationship between Q(t) and d1 and d2 can be expressed as: Q(t)=Z(|d1-d2|) Q(t) is negatively correlated with the size of |d1-d2|. When |d1-d2| is 0, it means that the vehicle is driving in the center of the tunnel. When |d1-d2| is larger, it means that the vehicle is closer to the sides of the tunnel. A(t) is positively correlated with the size of |d1-d2|. When |d1-d2| is 0, A(t) is the minimum value. The larger |d1-d2| is, the larger A(t) is.

7. The method for detecting lane departure or illegal lane change in a tunnel based on eye features according to claim 5, characterized in that: The changes in the driver's pupil area and the vehicle's lane departure or illegal lane change are as follows: When A(t) fluctuates within a certain range, and the maximum value of the fluctuation range is A max >1.2A0, minimum value A min When the value is less than 0.8A0, the vehicle is judged to have lane deviation. When A(t) increases first and then decreases, and its final value tends to the stable value A0, and this change process lasts for more than 3 seconds, it is judged that the vehicle is changing lanes illegally.

Images