Vehicle visibility shelter detection

Abstract

The invention provides vehicle visibility occlusion detection. A vehicle includes a window separating a cabin of the vehicle from an area external to the vehicle. The window includes an outer surface. The imaging device is in the cabin and is configured to capture an image of a region external to the vehicle through the window. A control circuit is in communication with the imaging device and is configured to receive an image from the imaging device, detect an occlusion in the image data corresponding to a substance on the window, and generate an output in response to detecting the occlusion, where the imaging device is disposed within 50 mm from the window.

Description

Technical Field

[0001] The present disclosure relates generally to image processing in a vehicular environment, and more particularly to vehicular visibility obstruction detection. Background Art

[0002] Conventional image processing techniques for detecting moisture conditions on or around a vehicle may be limited. Summary of the invention

[0003] According to a first aspect of the present disclosure, a vehicle includes a window that separates a cabin of the vehicle from an area outside the vehicle. The vehicle includes an imaging device in the cabin and configured to capture an image of the area outside the vehicle through the window. The vehicle includes a control circuit in communication with the imaging device. The control circuit is configured to: receive an image from the imaging device; process the image in a fully connected data description neural network (FCDDNN) trained to determine moisture content; detect an occluder in the image; associate the occluder with a substance on the window; classify the substance as water or debris based on moisture content; and generate an output in response to detecting the occluder.

[0004] Embodiments of the first aspect of the present disclosure may include any one or a combination of the following features:

[0005] - a wiper and a window cleaning system, the wiper being selectively movable along the window cleaning system, the window cleaning system including a controller in communication with the control circuit and configured to operate the wiper in response to the output;

[0006] - a vehicle window cleaning system including a pump for pressurizing a cleaning fluid for application to the vehicle window, wherein the controller is configured to selectively operate the pump in response to the output;

[0007] - the output generated by the control circuit is based on the classification of the obstruction as water or debris;

[0008] - the controller is configured to operate the wiper in response to classifying the matter as water and to operate the pump in response to classifying the matter as debris;

[0009] - The imaging device is set within 50 millimeters (mm) from the vehicle window;

[0010] - The control circuit includes a training module, which trains the FCDDNN based on manual operation feedback;

[0011] - manual operation feedback includes manual operation of at least one switch controlling at least one of a wiper and a pump;

[0012] - the training module is configured to train the FCDDNN to distinguish between a light source in an area outside the vehicle window and a substance on the vehicle window;

[0013] - the control circuit is configured to classify the optical distortion level using the moisture content of the substance;

[0014] - the vehicle window is a windshield of the vehicle extending at an oblique angle relative to the lens of the imaging device;

[0015] -The imaging device is set within 15 mm of the vehicle window;

[0016] - the control circuit is configured to detect another vehicle in front of the vehicle; and classify the material as tire splash from the other vehicle; and

[0017] - A control circuit wherein the automatic speed control system is configured to control the speed of the vehicle based on classifying the matter as tire splash from said other vehicle.

[0018] According to a second aspect of the present disclosure, a vehicle includes a windshield that separates a cabin of the vehicle from an area outside the vehicle. The vehicle includes an imaging device in the cabin and configured to capture an image of the area outside the vehicle through the windshield, and a control circuit that communicates with the imaging device. The control circuit is configured to receive an image from the imaging device, detect an obstruction in the image corresponding to a substance on the windshield, and generate an output in response to detecting the obstruction. The imaging device is disposed within 50 millimeters (mm) of the windshield.

[0019] Embodiments of the second aspect of the present disclosure may include any one or a combination of the following features:

[0020] - a wiper and window cleaning system, the wiper being selectively movable along the windshield, the window cleaning system including a controller in communication with the control circuit and configured to operate the wiper in response to the output;

[0021] - The imaging device is placed within 15 mm of the windshield; and

[0022] - The control circuit is configured to classify the optical distortion level according to the moisture content of the substance.

[0023] According to a third aspect of the present disclosure, a vehicle includes a windshield that separates a cabin of the vehicle from an area outside the vehicle. The vehicle includes a wiper that is selectively movable along the windshield. The vehicle includes a window cleaning system that includes a controller configured to operate the wiper. The vehicle includes an imaging device that is in the cabin and is configured to capture an image of an area outside the vehicle through the windshield. The imaging device is disposed within 50 millimeters (mm) of the windshield and includes a lens. The windshield extends at an oblique angle relative to the lens. The vehicle includes a control circuit that communicates with the imaging device and the window cleaning system. The control circuit is configured to: receive an image from the imaging device; process the image in a fully connected data description neural network (FCDDNN) trained to determine moisture content; detect an occlusion in the image; associate the occlusion with a substance on the windshield; classify the substance as water or debris based on the moisture content; and generate an output in response to classifying the substance as water. The controller is configured to operate the wiper in response to the output.

[0024] Embodiments of the third aspect of the present disclosure may include any one or a combination of the following features:

[0025] - A window cleaning system includes a pump for pressurizing cleaning fluid for application to the windshield, wherein the controller is configured to selectively operate the pump in response to classifying the material as debris.

[0026] These and other features, advantages and objectives of the present disclosure will be further understood and appreciated by those skilled in the art by referring to the following description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In the attached picture:

[0028] Figure 1 is a perspective view of a vehicle incorporating a moisture detection system according to one aspect of the present disclosure;

[0029] Figure 2 is a functional block diagram of a moisture detection system for a vehicle according to one aspect of the present disclosure;

[0030] Figure 3 is an exemplary cross-sectional view of a vehicle incorporating an imaging device within an interior of the vehicle adjacent a windshield of the vehicle for detecting moisture conditions in an area exterior to the vehicle;

[0031] Figure 4A is an exemplary image captured by an imaging device positioned in an interior of a vehicle away from a windshield such that water droplets on the windshield may be detected;

[0032] Figure 4Bis an exemplary image captured by an imaging device positioned within an interior of a vehicle and proximate a windshield, illustrating detection of water streaks on the windshield;

[0033] Figure 5 is a representation of a fully convolutional data description (FCDD) network that processes an image captured by an imaging device positioned proximate a windshield and produces image data showing occluders in the image;

[0034] FIG. 6A to FIG. 6C is a captured image and image data showing an obstruction on the windshield 18 of the vehicle after being processed by the FCDD network;

[0035] FIG. 7A to FIG. 7B is an exemplary image captured by an imaging device for a vehicle illustrating a splash event along a path of the vehicle, Figure 7B The splash zone covering the splash event is shown;

[0036] Figure 8 is an exemplary process performed by a moisture detection system according to one aspect of the present disclosure. DETAILED DESCRIPTION

[0037] Reference will now be made in detail to the present preferred embodiments of the present disclosure, examples of which are shown in the accompanying drawings. Wherever possible, the same reference numerals will be used throughout the drawings to refer to the same or similar parts. In the drawings, the depicted structural elements are not drawn to scale, and certain parts are enlarged relative to other parts for the purpose of emphasis and understanding.

[0038] As required, detailed embodiments of the present disclosure are disclosed herein; however, it should be understood that the disclosed embodiments are merely examples of the present disclosure, which may be implemented in various alternative forms. The drawings are not necessarily designed in detail; some schematic diagrams may be enlarged or minimized to show functional overviews. Therefore, the specific structural details and functional details disclosed herein should not be interpreted as limiting, but only as a representative basis for teaching those skilled in the art to adopt the present disclosure in different ways.

[0039] For the purpose of description herein, the terms "upper", "lower", "right", "left", "rear", "front", "vertical", "horizontal" and their derivatives shall refer to the following: Figure 1 The concepts of orientation in the drawings are presented in detail. However, it should be understood that the concepts can assume various alternative orientations unless expressly indicated to the contrary. It should also be understood that the specific devices and processes shown in the drawings and described in the following specification are merely exemplary embodiments of the inventive concepts defined in the appended claims. Therefore, unless the claims expressly state otherwise, specific dimensions and other physical characteristics related to the embodiments disclosed herein should not be considered limiting.

[0040] The embodiments shown herein reside primarily in a combination of method steps and apparatus components related to visibility obstruction detection of a vehicle 14. Accordingly, apparatus components and method steps have been represented, where appropriate, by conventional symbols in the drawings, and only those specific details relevant to understanding the embodiments of the present disclosure are shown so as not to obscure the present disclosure with details that would be readily apparent to one of ordinary skill in the art having the benefit of the description herein. In addition, like reference numerals in the specification and drawings represent like elements.

[0041] Overall reference Figures 1 to 8 , the moisture detection system 10 uses image processing to detect conditions in an area 12 outside the vehicle 14, such as water on the outer surface 16 of the windshield 18 of the vehicle 14 or water in front of or around the vehicle 14. In general, the present moisture detection system 10 can provide enhanced space utilization within the vehicle 14 by allowing the imaging device 20 of the moisture detection system 10 to be positioned near the windshield 18 or away from the windshield while still detecting and classifying obstructions on the windshield 18. In addition, the moisture detection system 10 can provide enhanced visibility by optimizing the cleaning and / or manipulation of the windshield 18 or suggesting maneuvering the vehicle 14 to a target position to enhance visibility and / or control. The moisture detection system 10 can also provide more accurate detection of weather conditions (such as rain, humidity, fog, or other moisture conditions in the area 12 outside the vehicle 14), thereby allowing enhanced responsiveness to other vehicle systems.

[0042] Reference now Figure 1 , the moisture detection system 10 for a vehicle 14 may incorporate an imaging device 20 positioned within a cabin 22 of the vehicle 14. For example, the imaging device 20 may be positioned within a passenger compartment 24 of the vehicle 14 adjacent to a windshield 18 of the vehicle 14 and oriented in a vehicle front orientation. At least one wiper 26 is positioned on the outer surface 16 of the windshield 18 for clearing water, debris, or other matter from the windshield 18. As will be described with respect to Figure 2 As further described, a window cleaning system may be employed to clean the windshield 18 or any of a plurality of windows of the vehicle 14 that may have accumulated moisture or other material on the window from an area 12 outside the vehicle 14. The window cleaning system may include a nozzle 28 positioned between a hood 30 of the vehicle 14 and the windshield 18 of the vehicle 14 adjacent to the wiper 26 and configured to spray a cleaning fluid onto the windshield 18. The cleaning fluid may include a detergent such as methanol, ethylene glycol, or other fluid for clearing material from the outer surface 16 of the windshield 18 when used with the wiper 26. It is contemplated that the cleaning fluid may include water, which in some examples may be heated to assist in the defrosting operation of the windshield 18.

[0043] Still reference Figure 1 , a plurality of distance sensors 34, 36, 38 are incorporated into the vehicle 14 to detect other objects in the area 12 outside the vehicle 14, such as other vehicles 14. The detection device may include ultrasonic and / or infrared detectors and a camera 38 (such as an imaging device) configured to detect the distance from the vehicle 14 to surrounding objects in the area 12 outside the vehicle 14. Thus, images from the imaging device 20 may be used to detect the distance from the vehicle 14 to objects within the field of view of the imaging device.

[0044] As about Figure 2 As further described, the distance sensors 34, 36, 38 may include any of a radio detection and ranging sensor (radar 34), a light detection and ranging sensor (lidar 36), and a camera 38. In some examples, an ultra-wideband (UWB) sensor is employed to detect objects in the exterior area 12. Generally speaking, the distance sensors 34, 36, 38 may be configured to detect intersection traffic events and / or detect objects in the blind spot of the vehicle 14 to assist the user of the vehicle 14 in maneuvering the vehicle 14. In this example, the data collected by the detection sensors is used by the moisture detection system 10 to allow the moisture detection system 10 to detect the distance (e.g., the following distance 120) from the vehicle 14 to the splash event 118 and to classify the importance, relevance, or priority of the splash event 118 based on the distance.

[0045] Still reference Figure 1 , the vehicle 14 includes a plurality of wheels 40, each wheel having a tire 42 that interacts with a driving surface 44 of the vehicle 14. The friction between each tire 42 and the driving surface 44 may be affected by the moisture condition between the tire 42 and the driving surface 44. Therefore, the moisture detection system 10 may be used to control the movement of the wheels 40 to enhance the maneuverability of the vehicle 14 along the driving surface 44.

[0046] The vehicle 14 includes at least one lighting assembly 46, such as a headlamp assembly having a headlamp 48 configured to illuminate an area 12 outside the vehicle 14. For example, the lighting assembly can be configured to illuminate the area 12 outside the vehicle 14 at multiple lighting levels (e.g., high beam, low beam, etc.). The moisture detection system 10 can enhance control of the lighting assembly and its power level and / or lighting level. For example, a moisture condition detected by the moisture detection system 10 in the exterior area 12 can cause the moisture detection system 10 to control the power level of the lighting assembly due to reduced visibility of the moisture condition.

[0047] Reference now Figure 2, the moisture detection system 10 includes an imaging system 50, a distance detection system 52, and a response control system 54 in communication with the imaging system 50 and the distance detection system 52. In general, data from the imaging system 50 and / or the distance detection system 52 is transmitted to the response control system 54, and the response control system 54 can control one or more visibility features and / or motion control features of the vehicle 14. For example, the response control system 54 includes one or more controllers 56, 72, which have at least one processor and a memory in communication with the processor. The memory can store instructions that, when executed by the processor of the response control system 54, cause the response control system 54 to perform various tasks related to enhancing the visibility of the area 12 outside the cabin of the vehicle 14 and / or the motion control of the vehicle 14. In general, one or more of the controllers 56, 72 and / or other electrical components that provide decision making via software can be referred to as control circuits.

[0048] For example, the response controller 56 of the response control system 54 may include a motion control unit 58 and a visibility control unit 60. The visibility control unit 60 may be configured to control a window cleaning system, a light control system 62, and / or any other system that affects visibility through one or more windows of the vehicle 14. The motion control unit 58 may control communications with one or more vehicle systems, such as a powertrain 64 of the vehicle 14, a brake system 66 of the vehicle 14, or any other motion control system for the vehicle 14. For example, the motion control unit 58 may include a speed sensor in the powertrain 64 that detects the rotation of a gear arrangement in the powertrain 64, however, other speed sensors may be used to detect or infer the speed of the vehicle (RF waves, inductive sensing, capacitive sensing, etc.). In addition, the response control system 54 may include and / or communicate with a display 68, such as a human-machine interface (HMI 70) in the cabin of the vehicle 14. The display 68 is configured to present messages to a user and / or allow a user to control the window cleaning system, the light control system 62, or any other aspect related to visibility and / or motion of the vehicle 14. Generally speaking, the response control system 54 may be configured to actively control visibility and / or motion characteristics of the vehicle 14 , or to passively present messages at the display 68 to indicate visibility and / or motion target operations for user initiation.

[0049] Continue to refer Figure 2 , the window cleaning system may include a window cleaning controller 72 that controls the operation of a pump 74 that pressurizes the cleaning fluid to spray the cleaning fluid onto the windshield 18 via the nozzle 28, as previously described. A valve 76 may fluidly intervene between the pump 74 and the nozzle 28 to selectively allow the cleaning fluid to enter the nozzle 28.

[0050] The window cleaning controller may also communicate with a motor 78 that drives the wiper 26. For example, the motor 78 may be configured to rotate the wiper 26 on the windshield 18 in response to a signal from the window cleaning controller 72. In some examples, the speed of the vehicle 14 as detected by the speed sensor may be compared to the detected splash event 122, and in response to the detection, the motor 78 may be energized to operate at a specific number of revolutions per minute (RPM). At least one switch 80 communicates with the window cleaning controller 72 and / or directly communicates with the motor 78, the pump 74, and / or the valve 76 to control the dispensing of the cleaning fluid and / or the driving of the wiper 26 via manual interaction. For example, the at least one switch 80 may include a first mechanism 82 that dispenses the cleaning fluid on the windshield 18 and a second mechanism 84 that controls the operation of the wiper 26. For example, the first mechanism 82 may be a knob that causes the cleaning fluid to be dispensed when pulled or pushed, and the second mechanism 84 may be a button or knob that causes the wiper 26 to move on the windshield 18. It is contemplated that the window cleaning controller 72 may be omitted in some examples, and the response control system 54 may directly control the window cleaning operation. In such examples, at least one of the switches 80 intervenes between the valve 76 and the nozzle 28 (e.g., the first mechanism 82), and at least one of the switches 80 intervenes between the motor 78 and the wiper 26 (e.g., the second mechanism 84). In either example, the user may manually control the dispensing of the cleaning fluid and / or the operation of the wiper 26, and such operations may also or alternatively be automatically controlled by the response control system 54. In another example, instructions for initiating wiping of the windshield 18 and / or cleaning of the windshield 18 are indicated at the HMI 70, and automatic control of the window cleaning system is omitted.

[0051] The visibility control unit 60 may also or alternatively be configured to control the lighting assembly of the vehicle 14 based on the moisture conditions detected by the imaging system 50, as previously described. For example, in the event that it is raining in the area 12 outside the vehicle 14, the headlights 48 may be automatically activated, or the response control system 54 may transmit instructions to present a message at the display 68 for the user to activate the headlights 48 in response to detecting moisture conditions in the area 12 outside the vehicle 14. Additionally, the brightness level (e.g., binary high / low beam or fine brightness control) may be actively or passively controlled by the response control system 54 (e.g., presenting a message at the display 68).

[0052] In general, the motion control unit 58 may control various systems of the vehicle 14 related to motion control, such as the drive of the wheels 40 (e.g., torque values), the brakes of the braking system 66 (e.g., traction control, anti-lock brakes (ABS)), the steering of the vehicle 14 (e.g., maneuvering along the driving surface 44), or any other motion control unit 58 for the vehicle 14. Similar to the operation of the visibility control unit 60, in addition to or in lieu of automatic control of the maneuverability of the vehicle 14, features related to motion control of the vehicle 14 may be presented at the display 68. For example, in an at least partially semi-autonomous control mode of the vehicle 14, the visibility control unit 60 may transmit instructions to the user at the display 68 to reduce speed, maneuver the vehicle 14 left or right, etc., and in the alternative, speed, maneuvering left or right, etc. may be controlled in response to detecting a water condition.

[0053] Continue to refer Figure 2 , the various responses transmitted by the response control system 54 can be based on outputs from one or both of the imaging system 50 and the distance detection system 52. With specific reference to the imaging system 50, the imaging device 20 captures one or more images (captured images 86) of the area 12 outside the vehicle 14, which are then processed by the image processor 88 to detect moisture conditions on or around the vehicle 14 and produce one or more output images 90. The image processor 88 is in communication with a memory 91 storing instructions that, when executed, cause the image processor 88 to detect water marks, water droplets 108, splash events 118, splash sources 116, or any other optical distortion related to moisture condition detection and / or occluder detection. A fully convolutional data description network (FCDDNN 92) is included within or in communication with the memory 91, which can be a neural network that segments the image data and detects optical distortions in the captured images 86. The FCDDNN 92 may be trained by a training module 94 of the imaging system 50, which is provided with sample images and / or historical image data that exhibit optical distortion (e.g., optical distortion caused by moisture in the image). The FCDDNN 92 is used to detect portions of the image that have distortion due to moisture conditions and to use the reference image as a training module. Figure 5 is described in further detail.

[0054] In addition to detecting moisture conditions, the image processor 88 is also configured to detect any other obstructions within the field of view of the imaging device 20. For example, the obstruction may be environmental debris, such as excrement, leaves, branches, twigs, or any other material that may adhere to or fall on the outer surface 16 of the windshield 18. Thus, the imaging system 50 can distinguish between debris and water. For example, the FCDDNN 92 can be trained to score obstructions with certain levels of opacity, light transmittance, light distortion, etc. For example, debris may be associated with opacity, while water and / or other moisture conditions may be associated with the bending of light through the obstruction.

[0055] In general, the response control system 54 can use the distance information from the distance detection system 52 and the moisture classification and detection from the image processing system to initiate a vehicle 14 response. For example, if moisture above a certain threshold (e.g., a low visibility threshold) is detected on the windshield 18, the imaging system 50 can transmit an output to the response control system 54 indicating the environmental conditions outside the area. For example, the imaging system 50 can determine that the external area 12 is raining, snowing, or otherwise precipitating, and transmit an indication to the response control system 54 to activate the wipers 26. The image processing system can further distinguish between various forms of precipitation and / or obstructions on the windshield 18 to allow for a personalized response initiated by the control system based on the environmental conditions. For example, and as will be described with respect to the aforementioned figures, the imaging system 50 can classify the type of obstruction as moisture-related or light-related, and the signal transmitted from the imaging system 50 of the response control system 54 can depend on the type of obstruction.

[0056] For example, the imaging system 50 may detect excrement on the windshield 18 and transmit an output of the response control system 54 to control the pump 74 of the window cleaning system to spray cleaning fluid onto the windshield 18. In another example, the imaging system 50 detects splashing spray on the windshield 18 from another vehicle 14 in front of the vehicle 14 and, in response, transmits a signal to the response control system 54 to control the HMI 70 or transmits a message to the HMI so that the user controls the vehicle 14 to slow down or adjust the positioning of the vehicle 14. The lighting assembly may also be controlled in response to detecting a moisture condition to illuminate an area 12 outside the vehicle 14 at a specific power level / lighting level and / or turn on or off the headlights 48. These responses are exemplary and non-limiting, such that the moisture detection system 10 may perform any combination of responses in response to the classification of the imaging system 50.

[0057] Still reference Figure 2, the response control system 54 can monitor feedback from the window cleaning system, the moisture control system, the HMI 70, or other vehicle systems to optimize the response of the moisture detection system 10. For example, when the imaging system 50 classifies the environmental conditions outside the area, the imaging system 50 can transmit an instruction or a signal that there is light rain in the area 12 outside the vehicle 14 to the response control system 54. In response to detecting light rain, the response control system 54 can initiate wiper control at a low speed. However, manually controlling the wipers 26 via the second switch 80 to stop wiping the windshield 18 can end the automatic control of the wipers 26. The response control can detect feedback (e.g., the user manually turns off the wipers 26) and update or optimize future responses when light rain is detected. In this way, the detection and classification of environmental conditions (including moisture conditions) and the response to them can be enhanced to promote an optimized response by the user. It is contemplated that other examples related to other environmental conditions can be similarly optimized using feedback (e.g., manual adjustment of the lighting assembly, manual steering adjustment, braking adjustment, etc.). In some examples, user dismissal of a message presented at the HMI 70 may indicate that the response indicated in the message is undesirable.Other examples of feedback and control will be described with respect to the preceding figures.

[0058] Reference now Figure 3 , an exemplary arrangement of the imaging device 20 relative to the windshield 18 is shown to demonstrate the proximity of the imaging device 20 relative to the windshield 18 and the enhanced optical and spatial properties of the arrangement. Although shown as being disposed adjacent to the upper portion 96 of the windshield 18, it is contemplated that the imaging device 20 may be disposed along any portion of any window of the vehicle 14, allowing the imaging device 20 to capture a view of the area 12 external to the vehicle 14. In general, the software employed by the moisture detection system 10 of the present disclosure allows the positioning of the imaging device 20 to be at least partially independent of the distance between the lens 98 of the imaging device 20 and the vehicle window. For example, and as will be described with respect to Figure 4A and Figure 4BAs described, water (e.g., condensation, liquid water, frost) on the windshield 18 may appear different in an image depending on the distance of the imaging device 20 relative to the windshield 18. For example, when the imaging device 20 is spaced farther from the windshield 18 or very close to the windshield 18 (e.g., between 0 mm and 35 mm) relative to against the windshield 18, the water droplets 108 may appear as discrete circles, dots, or other geometric shapes that are more clearly defined along the edges 110 of the shapes. The processes and methods implemented by the moisture detection system 10 of the present disclosure may allow for the detection of moisture conditions at either or both of the first position 102 of the imaging system 50 (e.g., at a first distance) and the second position 104 of the imaging device 20 (e.g., at a second distance less than the first distance). For example, when the imaging device 20 is at the first distance from the windshield 18, the moisture on the windshield 18 may appear as a blurred water streak or less clear optical distortion because the imaging device 20 is closer to the water on the windshield 18 than when positioned at the second distance. By providing a distance independent arrangement, the moisture detection system 10 allows for enhanced separation within the cabin, a greater and more detailed field of view for the imaging device, and by providing a universal application to allow the imaging device to be located at different distances from the windshield 18 .

[0059] Continue to refer Figure 3 , the windshield 18 may extend at an inclined angle 106 relative to a vertical and / or horizontal orientation. For example, the windshield 18 may be tilted upward relative to the driving surface 44. The tilt angle 106 of the windshield 18 may further distort or otherwise affect the representation of moisture conditions in images captured by the imaging device. In examples where the imaging device 20 is in the second position 104, the distortion may be more apparent relative to being in the first position 102. Therefore, the challenges of capturing moisture conditions of the environment outside the vehicle 14 from images captured by the imaging device 20 inside the cabin may be greater than or different from the image processing used to detect moisture in images captured by the imaging device outside the cabin. For example, when falling on the windshield 18, a water drop 108 on the windshield 18 may expand to be longer than a water drop 108 on a vertical or more inclined surface (such as a passenger window, a rear window, an exterior camera cover, etc.).

[0060] Reference now Figure 4A and Figure 4B , showing the Figure 3 The difference between the image 86 captured by the first position 102 of the imaging device 20 relative to the second position 104. Figure 4A As shown (first position 102), rain on the outer surface 16 of the windshield 18 may be due to Figure 4B18 (second location 104) as droplets 108 having more clearly defined discrete shapes or edges 110 than blur 112. Using the current algorithms and processes of the moisture detection system 10, such optical distortions, albeit slight, may also be detected. For example, when the imaging device 20 is positioned away from the windshield 18, the droplet detection algorithm may be suitable for detecting moisture conditions, and when the imaging device 20 is positioned closer to the windshield 18 ( Figure 4B ), the imaging system 50 can employ a blur 112 detection algorithm to detect moisture conditions on the windshield 18. Thus, the moisture detection system 10 can provide enhanced separation within the vehicle 14 by providing the imaging device 20 proximate to the windshield 18.

[0061] In some examples, the second position 104 is within 50 mm of the windshield 18. In some examples, the second position 104 is between 0 mm and 35 mm from the windshield 18. In some examples, the second position 104 is positioned between 0 mm and 10 mm from the windshield 18. In each of these examples, the second position 104 is close to the windshield 18 to provide enhanced visibility and spacing within the cabin. In these examples, the first position 102 is farther from the windshield 18 than the second position 104.

[0062] Reference now Figure 5 , showing the processing Figure 4B FCDDNN 92 is a deep learning model designed for unsupervised irregularity detection in data and operates by learning the underlying patterns and structures within the image data to distinguish normal patterns from abnormal patterns. FCDD may include a plurality of layers 114, each of which is a convolutional layer. Because each layer may be convolutional, FCDDNN 92 may capture spatial dependencies and maintain the spatial structure of the input data.

[0063] In the FCDDN, input data is passed through a series of convolutional layers 114, which extract relevant features at different levels of abstraction. These convolutional layers 114 may be followed by pooling layers 114 to downsample the feature map and reduce the spatial dimension of the feature map. The output of the convolutional layer 114 may then be flattened and fed into a fully connected layer 114, which performs further feature extraction and maps the learned features to irregularity scores. Irregularity detection may be achieved by comparing the calculated irregularity score with a predefined threshold, where a score above the threshold indicates an abnormal instance. In this example, the predefined threshold may correspond to edge continuity of shapes in the image to detect distortion (e.g., blur) caused by moisture. These thresholds may be actively adjusted based on the user feedback previously described (e.g., the user manually activates the wiper 26 and / or the cleaning process, the user manually activates the headlight 48, etc.).

[0064] Reference now FIG. 6A to FIG. 6C , depicts an exemplary captured image 86 from the imaging device 20 directed toward the windshield 18 and filtered image data (output image 90) indicating moisture conditions on the windshield 18. The imaging system 50 may process the captured image 86 in an image processor 88, including processing the captured image 86 in one or more machine learning models and / or neural networks. The imaging system 50 may employ edge detection techniques, histogram equalization, linear filters, image segmentation, convolution, or any combination of any image processing techniques to detect at least one portion of the captured image having distortion or occlusion. For example, and with reference to Fig. 6A and Figure 6B , the imaging system 50 can detect blur 112 on the windshield 18 corresponding to a wet condition (e.g., moisture on the windshield 18). In another example, the imaging system 50 detects a non-moisture condition, such as one of the wipers 26 moving across the windshield 18 ( Figure 6C ). In these examples, detection may be applied to the captured image 86 to correlate groups of pixels of the captured image 86 with surrounding pixels in order to classify the image data as moisture-related or non-moisture-related. For example, the imaging system 50 may employ lane detection to identify one or more lane lines defining the plurality of lanes 126. In general, object classification may be performed by the imaging system 50, such as classifying other vehicles 14, street lights, trees, or any other object captured in the captured image 86.

[0065] Generally speaking, previous Figure 5The FCDDNN 92 described can be used to provide a moisture estimate. Based on the moisture estimate, which can be a continuously varying value or a binary wet or non-wet value, the response control system 54 can determine the response of the moisture detection system 10. By combining the various layers 114 of the FCDDNN 92 with manual feedback from the user, as previously described with respect to Figure 2 As described, the techniques employed by the moisture detection system 10 can provide enhanced responses. For example, based on the training of the FCDDNN 92, the wipers 26 can be activated when or before the user would normally activate the wipers 26. In addition, the speed of the wipers 26 and / or the activation of the pump 74 for cleaning the windshield 18 can be optimized based on the moisture content and / or the previous activation of the wipers 26 or the pump 74. The headlights 48 of the vehicle 14 can also or alternatively be activated with limited false activations. For example, the FCDDNN 92 detects image data and classifies the image data as oncoming headlights 48, and distinguishes oncoming headlights 48 from moisture conditions or non-moisture conditions. In addition, the FCDDNN 92 can distinguish objects with significant moire effects (e.g., fences). In some examples, manual feedback includes manual operation of the wipers (e.g., via at least one switch 80) and / or manipulation of the vehicle 14 away from the splash event 118. Additionally, and as will be described below, adjusting the vehicle 14 to another lane or maintaining position in the current lane may be manual feedback (eg, ignoring a suggestion to change lanes or manually overriding to move into another lane).

[0066] Reference now Fig. 7A and Figure 7B , identification of at least one splash source 116 of a splash event 118 may be performed by the imaging system 50 of the moisture detection system 10. For example, the splash source 116 may include a vehicle 14 and a non-vehicle 14. The splash source 116 may be a tire 42, a vehicle body, a bridge, an overpass, construction equipment, a fire hydrant, or any other source. Thus, the imaging system 50 may be configured to classify the source using any of the previous image detection techniques previously described.

[0067] Continue to refer Fig. 7A and Figure 7B, the imaging system 50 may also or alternatively determine the position of the splash source 116 relative to the vehicle 14. Based on the position of the splash source 116 (e.g., source lane 126a), the response control system 54 may control any of the window cleaning system, the lighting system, the HMI 70, and / or the motion control system. For example, the imaging system 50 may locate the splash source 116 in another lane of the plurality of lanes 126 other than the lane of the vehicle 14, and based on the position of the splash source 116 in the other lane, a driving route may be recommended to maintain the current lane of the vehicle 14. Therefore, in general, the moisture detection system 10 may control or indicate a message to control the vehicle 14 to adjust the lane based on the position of the splash source 116. In an example, as Figure 7B As shown, the imaging system 50 detects a plurality of splash sources 116, and the response control system 54 can recommend one of a plurality of lanes 126 based on the proximity of the vehicle 14 to one or more of the plurality of splash events 118 (e.g., based on the following distance 120 of the vehicle 14 to other vehicles 14, based on the relative positions of other vehicles 14 to other splash sources (such as potholes)). In other words, the moisture detection system 10 can detect a first splash source 116 and a second splash source 116, classify each splash source 116 by a correlation value, and recommend a lane or other position of the vehicle 14 based on the correlation value of the first splash source and the second splash source 116. As will be described below, the moisture detection system 10 can also recommend a target following distance, a target speed, etc. in response to the correlation value.

[0068] Specific reference Figure 7B , at least one splash zone 122 may be determined or calculated by the imaging system 50 based on the image data from the initial image. For example, any of the prior techniques described (including the application of the FCDDNN 92) may allow the imaging system 50 to estimate the size of the splash zone 122, such as the depth D, width W, and height H of the splash zone. The size of the splash zone 122 may be determined based on the following distance 120 from the splash source to the vehicle 14 and the captured image 86. For example, the previously described techniques may be used to estimate the size of the splash zone 122, such as the depth D, width W, and height H of the splash zone. Figure 1 and Figure 2 The depicted distance sensors 34, 36, 38 determine the following distance 120 and may detect the forward-most location of the splash based on image processing of the captured image 86. The difference between the following distance 120 and the forward end of the splash zone 122 may be calculated by the control circuitry to determine the depth D of the splash zone 122. In this manner, the distance detection and position detection of the system 10 may determine the origin of the splash zone, the size of the splash zone, the likelihood of contacting water in the splash zone based on distance and / or velocity, and the like.

[0069] In this example, the imaging system 50 detects three splash zones 122, where each of the splash zones 122 has a different size. The size of the splash zones 122 may be caused by the size of the tire 42, the speed of the vehicle 14, the lane in which the vehicle 14 is located (e.g., road roughness, potholes, etc.), or any other factor that may affect the size of tire splash. For example, an uneven road that may cause puddles, bends in the road (e.g., slopes, turns in the road), etc. may further affect the size of the splash zone. Based on the size of the splash zone, the imaging system 50 may determine a following distance 120 from each or any of the splash sources 116 (e.g., other vehicles 14). In general, the size of the splash zone, the density of the splash zone, the duration of the splash zone, or any other aspect related to the magnitude of the splash zone can be detected by the imaging system 50 and used to classify the priority or ranking (e.g., relevance value) of the splash event 118 to determine a target lane among the plurality of lanes 126, a target following distance 120 (e.g., a minimum following distance) for the vehicle 14 relative to other vehicles 14, activation of a window cleaning system, presentation of a message at the HMI 70, control of the vehicle 14, or any of the responses previously described. The direction of the splash zone (which may be affected by wind speed and direction) can also be detected to enhance the estimation of the splash zone 122 and further enhance the response determined by the response control system 54. Such wind speed and direction can be detected from a weather sensor or a global positioning system (GPS) in communication with the moisture detection system 10.

[0070] The moisture detection system 10 may further or alternatively estimate the density of moisture conditions, such as the density of the splash zone 122 or the density of rain. For example, based on the amount, distribution, or pattern of haze 112 or other moisture spots detected on the windshield 18, the moisture detection system 10 may estimate the density of rain and activate the wipers 26 or transmit an instruction to the user to operate the wipers 26 in response to the amount, distribution, or pattern exceeding a threshold or matching a target distribution or pattern. For example, if more than 25%, 50%, or 90% of the windshield 18 has haze 112 in the field of view of the imaging device 20, the imaging system 50 may transmit a signal to the response control system 54 to activate the wipers 26. In other examples, the size of the water droplets 108 and / or water droplets 112 may be classified by the FCDDNN 92 and compared to stored moisture conditions to determine the response of the moisture detection system 10.

[0071] In some examples, the moisture detection system 10 can access the location of the vehicle 14 relative to the type of road. For example, GPS can provide the location of the vehicle 14 and thus the type of road (e.g., highway, city driving, etc.). In this way, the number of lanes, direction of movement, construction, and / or mode of operation of the road (e.g., one-way road, interstate road with a median, highway without a median) can be determined. Therefore, the moisture detection system 10 can give some lanes priority over other lanes, or further adjust the following distance 120 based on the type of road. For example, because the typical speed of the vehicle 14 on the interstate is faster than on another road, the moisture detection system 10 can suggest a "slow lane" of two lanes based on the moisture conditions detected on the interstate.

[0072] Although the source lane 126a in the plurality of lanes 126 is shown as having the same traffic direction as another lane in the plurality of lanes 126 in the previously described features, in some examples, the plurality of lanes 126 includes a lane with a first traffic direction and a lane with a second traffic direction. In this example, the imaging system 50 is configured to classify the splash source 116 as being in the lane with the first traffic direction and / or to classify the splash source 116 as being in the lane with the second traffic direction. For example, the imaging system 50 may detect an oncoming headlamp 48 in an adjacent lane and, in response to the detection, classify the adjacent lane as having an opposite traffic direction. Thus, splash from oncoming traffic may be compared to splash from the traffic ahead, and the moisture detection system 10 may provide an enhanced response determination of the target lane of the vehicle 14 based on either or both of the splash events 118 in the first traffic direction and the second traffic direction. It is also contemplated that the distance detection system 52 described previously may be used to further determine the target lane of the vehicle 14 by tracking the following distance 120 from the vehicle 14 to the vehicle ahead 14.

[0073] Reference now Figure 8, an exemplary process 800 performed by the moisture detection system 10 is shown for use with splash detection and moisture condition detection on the windshield 18. At step 802, the imaging device 20 captures an image of the area 12 outside the vehicle 14. At step 804, the image processor 88 detects an event in the area immediately inside the vehicle 14. For example, the event can be a moisture condition, such as a splash event 118, water on the outer surface 16 of the windshield 18, an object on the windshield 18, or any other event associated with visibility obstruction or distortion. At step 806, the event is classified as being associated with the windshield 18 (e.g., on the windshield 18) or spaced from the windshield 18. For example, if the event is a splash event 118 associated with other vehicles 14 in front of the vehicle 14, rather than water splashing on the windshield 18 itself. It is contemplated that the splash event 118 can be classified as both an event on the windshield 18 and an event outside the windshield 18, depending on the following distance 120 between the splash source 116 and the vehicle 14. If the event is an obstruction on the exterior surface 16 of the windshield 18, the image processor 88 may classify the obstruction at step 808. For example, the imaging system 50 may classify the obstruction as debris or water. Based on the classification of the obstruction, at step 810, the moisture detection system 10 determines a response. For example, in the case where the obstruction is debris (such as wet debris), the response may be to initiate the application of a cleaning fluid on the windshield 18, or in the case of a moisture condition, the response may be to activate one or more of the wipers 26 on the windshield 18 or adjust their speed. At step 812, an output is transmitted to initiate the response determined at step 810. It is contemplated that the response may alternatively be a more passive response, such as presenting a message at the display 68 to instruct the user to implement one or more of the functions that may alternatively be automatically implemented.

[0074] At step 814, feedback in the form of manual adjustments or no action (e.g., the user did not follow the recommendation) is transmitted to the response control system 54 and / or the imaging system 50 to further refine the response determined in future events. For example, if the user is instructed to activate the wipers 26, and the user does not activate the wipers 26, the imaging system 50 may increase the target moisture content used to determine wiper activation to a threshold to limit erroneous responses to future calculations. Such a threshold may be the previously described threshold of the FCDDNN 92 or another threshold.

[0075] If the event is not related to the condition of the windshield 18, as determined in step 806, the process can continue to determine the location of the splash event 118 at step 816. For example, using the captured image 86, the imaging system 50 can detect the source and / or location (e.g., splash lane 126a) of the splash event 118 (e.g., tire splash) caused around the vehicle 14. At step 818, a response is determined by the response control system 54. For example, the response can be to adjust the following distance 120 between the splash source 116 and the vehicle 14 by reducing the speed of the vehicle 14. In other examples, the response includes maneuvering the vehicle 14 to another lane in the plurality of lanes 126. In other examples, the response includes presenting instructions at the HMI 70 via messaging to instruct the user to maneuver the vehicle 14, adjust the speed of the vehicle 14, etc. Other examples of responses include adjustments to the window cleaning system, such as activation of the wipers 26, adjustment of the speed of the wipers 26, activation of the pump 74 for applying cleaning fluid, etc. At step 820, an output can be transmitted to initiate the response. Similar to step 812 , at step 822 , feedback in the form of the user's action or inaction to overrule the response transmitted by the response control system 54 is returned to the moisture detection system 10 for use in enhancing response determinations in future situations where a splash event 118 is detected.

[0076] Generally speaking, the present moisture detection system 10 enhances the response of the vehicle 14 to limit visibility obstructions and / or distortions of the area 12 outside the vehicle 14. The image processing techniques employed by the moisture detection system 10 can enhance the space within the interior of the vehicle 14 by allowing the imaging device 20 to be positioned close to the windshield 18. In addition, the image processing techniques employed herein can enhance the detection of splash sources 116 and / or moisture conditions on the outer surface 16 of the windshield 18. Based on the detection of these moisture events, the moisture detection system 10 can provide a fast response time to clean the windshield 18 and / or optimize the maneuvering of the vehicle 14.

[0077] As used herein, the term "and / or" when used with a list of two or more items means that any one of the listed items may be employed alone, or any combination of two or more of the listed items may be employed. For example, if a composition is described as containing components A, B, and / or C, the composition may contain: A alone; B alone; C alone; a combination of A and B; a combination of A and C; a combination of B and C; or a combination of A, B, and C.

[0078] In this document, relational terms such as first and second, top and bottom, etc. are used solely to distinguish one entity or action from another entity or action, without necessarily requiring or implying any actual such relationship or order between such entities or actions. The terms "comprises," "comprising," or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element preceded by "comprises ..." does not, in the absence of more constraints, preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0079] As used herein, the term "about" means that the amount, size, formula, parameter and other quantity and characteristics are not exact, nor need to be exact, but may be approximate and / or larger or smaller as needed to reflect tolerances, conversion factors, rounding, measurement errors, etc. and other factors known to those skilled in the art. When the term "about" is used to describe the endpoints of a value or range, the disclosure should be understood to include the specific value or endpoint mentioned. Whether or not the endpoints of a numerical value or range in this specification are described as "about", the endpoints of the numerical value or range are intended to include two embodiments: one modified by "about" and one not modified by "about". It should also be understood that the endpoints of each of the ranges are significant in relation to the other endpoint and independently of the other endpoint.

[0080] As used herein, the terms "substantially," "substantially," and variations thereof are intended to indicate that the feature being described is equal to or approximately equal to a value or description. For example, a "substantially planar" surface is intended to indicate a planar or approximately planar surface. Additionally, "substantially" is intended to indicate that two values ​​are equal or approximately equal. In some embodiments, "substantially" may indicate that values ​​are within about 10% of each other, such as within about 5% of each other, or within about 2% of each other.

[0081] Unless expressly indicated to the contrary, as used herein, the terms "the," "a," or "an" mean "at least one" and should not be limited to "only one." Thus, for example, reference to "a component" includes embodiments having two or more such components unless the context clearly indicates otherwise.

[0082] It should be understood that changes and modifications may be made to the foregoing constructions without departing from the concepts of the present disclosure, and it should be further understood that such concepts are intended to be encompassed by the appended claims unless the claims expressly state otherwise by their language.

[0083] According to the present invention, a vehicle is provided, the vehicle having: a window that separates a cabin of the vehicle from an area outside the vehicle; an imaging device that is in the cabin and is configured to capture an image of the area outside the vehicle through the window; a control circuit that communicates with the imaging device, the control circuit being configured to: receive an image from the imaging device; process the image in a fully connected data description neural network (FCDDNN) trained to determine moisture content; detect an occlusion in the image; associate the occlusion with a substance on the window; classify the substance as water or debris based on the moisture content; and generate an output in response to detecting the occlusion.

[0084] According to one embodiment, the invention also features a wiper selectively movable along a vehicle window and a window cleaning system including a controller in communication with the control circuit and configured to operate the wiper in response to the output.

[0085] According to one embodiment, a vehicle window cleaning system includes a pump for pressurizing a cleaning fluid for application to a vehicle window, wherein a controller is configured to selectively operate the pump in response to the output.

[0086] According to one embodiment, the output generated by the control circuit is based on classifying the obstruction as water or debris.

[0087] According to one embodiment, the controller is configured to operate the wipers in response to classifying the matter as water, and to operate the pump in response to classifying the matter as debris.

[0088] According to one embodiment, the imaging device is positioned within 50 millimeters (mm) of the vehicle window.

[0089] According to one embodiment, the control circuit includes a training module that trains the FCDDNN based on manual operation feedback.

[0090] According to one embodiment, the manual operation feedback comprises manual operation of at least one switch controlling at least one of a wiper and a pump.

[0091] According to one embodiment, the training module is configured to train the FCDDNN to distinguish between light sources in an area outside the vehicle window and matter on the vehicle window.

[0092] According to one embodiment, the control circuit is configured to classify the optical distortion level by moisture content of the substance.

[0093] According to one embodiment, the vehicle window is a windshield of the vehicle extending at an oblique angle relative to the lens of the imaging device.

[0094] According to one embodiment, the imaging device is arranged within 15 mm of the vehicle window.

[0095] According to one embodiment, the control circuit is configured to: detect another vehicle in front of the vehicle; and classify the substance as tire splash from the other vehicle.

[0096] According to one embodiment, the invention also features an automatic speed control system for a vehicle in communication with the control circuit, wherein the automatic speed control system is configured to control a speed of the vehicle based on classifying the substance as tire splash from another vehicle.

[0097] According to the present invention, a vehicle is provided, the vehicle having: a windshield, the windshield separating the vehicle's cabin from an area outside the vehicle; an imaging device, the imaging device in the cabin and configured to capture an image of the area outside the vehicle through the windshield; and a control circuit, the control circuit communicating with the imaging device and configured to: receive an image from the imaging device; detect an obstruction in the image corresponding to a substance on the windshield; and generate an output in response to detecting the obstruction, wherein the imaging device is disposed within 50 millimeters (mm) of the windshield.

[0098] According to one embodiment, the invention also features a wiper that is selectively movable along the windshield and a window cleaning system that includes a controller in communication with the control circuit and configured to operate the wiper in response to the output.

[0099] According to one embodiment, the imaging device is arranged within 15 mm of the vehicle window.

[0100] According to one embodiment, the control circuit is configured to classify the optical distortion level by moisture content of the substance.

[0101] According to the present invention, a vehicle is provided, the vehicle having: a windshield that separates a cabin of the vehicle from an area outside the vehicle; a wiper that selectively moves along the windshield; a window cleaning system, the window cleaning system including a controller that is configured to operate the wiper; an imaging device that is in the cabin and is configured to capture an image of an area outside the vehicle through the windshield, wherein the imaging device is disposed within 50 millimeters (mm) of the windshield; and includes a lens, wherein the windshield extends at an oblique angle relative to the lens; and a control circuit that communicates with the imaging device and the window cleaning system, the control circuit being configured to: receive an image from the imaging device; process the image in a fully connected data description neural network (FCDDNN) trained to determine moisture content; detect an occlusion in the image; associate the occlusion with a substance on the windshield; classify the substance as water or debris based on the moisture content; and generate an output in response to classifying the substance as water, wherein the controller is configured to operate the wiper in response to the output.

[0102] According to one embodiment, a window cleaning system includes a pump for pressurizing cleaning fluid for application to a windshield, wherein a controller is configured to selectively operate the pump in response to classifying the matter as debris.

Claims

1. A vehicle comprising: a vehicle window, the vehicle window separating a cabin of the vehicle from an area outside the vehicle; an imaging device in the cabin and configured to capture an image of the area outside the vehicle through the window; a control circuit, the control circuit being in communication with the imaging device, the control circuit being configured to: receiving the image from the imaging device; processing the image in a fully connected data description neural network (FCDDNN) trained to determine moisture content; Detecting occluders in the image; Associating the obstruction with a substance on the vehicle window; classifying the matter as water or debris based on the moisture content; and An output is generated in response to detecting the occluder.

2. The vehicle of claim 1, further comprising: a windshield wiper selectively movable along the vehicle window; as well as A vehicle window cleaning system includes a controller in communication with the control circuit and configured to operate the wiper in response to the output. 3 . The vehicle of claim 2 , wherein the window cleaning system includes a pump for pressurizing cleaning fluid for application to the vehicle window, wherein the controller is configured to selectively operate the pump in response to the output.

4. The vehicle of claim 3, wherein the output generated by the control circuit is based on classifying the obstruction as water or debris. 5 . The vehicle of claim 4 , wherein the controller is configured to operate the wipers in response to classifying the matter as water and to operate the pump in response to classifying the matter as debris.

6. The vehicle of any one of claims 1 to 5, wherein the imaging device is disposed within 50 millimeters (mm) of the vehicle window.

7. The vehicle of claim 3, wherein the control circuit includes a training module that trains the FCDDNN based on manual operation feedback.

8. The vehicle of claim 7, wherein the manual operation feedback comprises manual operation of at least one switch that controls at least one of the wiper and the pump.

9. The vehicle of claim 7 or 8, wherein the training module is configured to train the FCDDNN to distinguish between light sources in the area outside the vehicle window and matter on the vehicle window.

10. The vehicle of any one of claims 1 to 9, wherein the control circuit is configured to classify the optical distortion level by the moisture content of the substance.

11. The vehicle according to any one of claims 1 to 10, wherein the vehicle window is a windshield of the vehicle extending at an oblique angle relative to the lens of the imaging device.

12. The vehicle of any one of claims 1 to 11, wherein the imaging device is arranged within 15 mm of the vehicle window.

13. The vehicle of any one of claims 1 to 12, wherein the control circuit is configured to: detecting another vehicle ahead of the vehicle; and The material is classified as tire splash from the other vehicle.

14. The vehicle of any one of claims 1 to 13, further comprising: An automatic speed control system for the vehicle in communication with the control circuit, wherein the automatic speed control system is configured to control the speed of the vehicle based on classifying the substance as tire splash from the other vehicle.

Images