02. Camera Flashcards
What is the main purpose of cameras in autonomous vehicles?
Cameras in autonomous vehicles serve to: enhance visibility, provide driver monitoring, and supply inputs for object and event detection and response systems.
Where are cameras typically positioned in autonomous vehicles?
Cameras are typically positioned to provide: front view (forward-facing) and mirrors view (side-facing).
What are the parking-related applications of cameras for enhancing visibility?
Parking-related applications include: parking assist, cross traffic alert, backover protection, and trailer steering assist.
What are the safety-related applications of camera systems?
Safety applications include: blind spot detection, rear collision warning, emergency braking, automatic cruise control, and pedestrian and vehicle detection.
What road features can cameras detect for autonomous driving?
Cameras can detect: lane markings (for lane detection and keeping), traffic signs (for recognition), and can assist with headlight control, bicycle recognition, and navigation at intersections.
What are the main applications of driver monitoring systems?
Driver monitoring applications include: drowsiness detection, distraction detection, optimization of airbag deployment, seat belt adjustment, and driver authentication.
What is detection range in the context of automotive cameras?
Detection range is the distance at which a camera can detect an object. It must be greater than or equal to the sum of reaction distance and braking distance.
What is the formula to calculate the required detection range?
The detection range (d) must satisfy: d ≥ Reaction Distance + Braking Distance, where:
- Reaction Distance [m] = Speed [km/h] × Reaction Time [s] × (1/3.6)
- Braking Distance [m] = Speed²[km/h]/(250 × f), where f depends on road conditions (0.8 for dry asphalt, 0.5 for wet asphalt, 0.3 for snow, 0.2 for ice).
What is angular resolution in camera systems?
Angular resolution is the ability to distinguish details of an object. For object detection, a minimum number of pixels is required at a distance equal to or greater than the vehicle’s stopping distance.
What is the formula for angular resolution?
Angular resolution: resβ(d) = n/arctan(wo/d), where:
- d: object distance (stopping distance)
- wo: object width (0.5m for short side, 1m for long side)
- n: required number of pixels for object detection (3px for short side).
What is Field of View (FOV) and why is it important?
Field of View (FOV) is the extent of the observable world seen at any given moment. It should be wide enough to cover the lane in front of the vehicle and adjacent lanes, ensuring detection of road lines, objects, and people even at short distances.
What are the two scenarios to consider when determining Field of View?
The two scenarios are: straight road and curved road. Each requires different FOV calculations based on visibility requirements.
How is FOV calculated for a curved road?
For a curved road: FOV = 2β(v) = (360° × d(v))/(2 × r(v) × π), where:
- FOV: field of view
- d(v): object distance
- r(v): curve radius
- v: vehicle speed
Note: According to ISO/FDIS 15623, 2002, only curve radiuses of minimum 250 m are typically considered.
What is frame rate in camera systems?
Frame rate is the number of consecutive images that can be taken in one second, typically ranging from 24 to 60 frames per second (fps) for automotive applications.
What is focal length and how does it relate to Field of View?
Focal length is the distance between the lens and the image sensor when the subject is in focus. It relates to FOV and sensor dimensions by:
- FOVα = 2arctan(hsensor/2f)
- FOVβ = 2arctan(wsensor/2f)
Where hsensor and wsensor are the sensor height and width, and f is the focal length.
In camera system calculations, what relationship is used to determine the image size of an object?
The relationship is: S/d = I/f, where:
- S: object size
- d: distance to object
- I: image size on sensor
- f: focal length
This derives from the principle of similar triangles in optics.
When selecting a camera for an autonomous braking system, what parameters must be considered?
Key parameters include: sensor resolution (pixels), physical sensor dimensions, pixel size, focal length, minimum object size to detect, and the required detection range based on vehicle speed and braking capabilities.
How is the minimum number of pixels needed to represent an object for detection calculated?
The minimum number of pixels depends on the object detection algorithm requirements. For reliable detection, objects typically need to occupy at least 20×20 pixels (or sometimes just 3 pixels on the shortest side in basic applications).
How is the maximum safe speed for an emergency braking system determined based on camera parameters?
It requires:
- Calculating the maximum distance at which the camera can reliably detect the minimum-sized object (using the S/d = I/f relationship)
- Subtracting the distance traveled during the system’s reaction time
- Calculating the speed that would allow the vehicle to stop within the remaining distance
- Considering the friction coefficient of the road surface.
