Question: How best can camera attributes be recorded to enable the location of observed objects to be calculated over time?
Background
Previous discussions in W3C Spatial Data on the Web (w3c/sdw#1137) and OGC GeoPose (opengeospatial/GeoPose#2) have helped to identify suitable attributes for image characteristics and camera orientation which are aligned with the proposed GeoPose standard.
Use Cases
Dashcam. A forward-facing camera mounted on a vehicle calculates location from a global navigation satellite system (GNSS) and heading from a compass, and these data can be captured in timed video metadata. Orientation only needs to be calculated in 2D as the vehicle is on the ground and low precision is sufficient as the camera has a wide field of view. Multiple cameras can be added to include concurrent rear and side views that may be beneficial to a human driver and which enable an autonomous vehicle to monitor the movement of nearby moving and static objects such as other road users, pedestrians and obstacles to avoid collision.
Drone with gimballed camera. An unmanned aerial vehicle (drone) calculates location from GNSS, height from an altimeter, orientation from a compass and gyro, and camera orientation from the gimbal controller. 3D location and orientation are required as the camera is airborne and more precision is needed due its zoom capability which can affect the field of view.
CCTV. A closed circuit television (CCTV) system includes multiple static cameras which are either fixed or have pan-tilt-zoom (PTZ) capabilities such as roadside traffic cameras. Moving objects can be sequentially tracked by multiple cameras or concurrently observed from multiple angles where fields of view overlap.
Requirements
The location and orientation should enable calculation of the camera's viewpoint and allow the location of observed objects to be estimated.
Attribute
Description
Identifier
Unique identifier for the camera that captured the image
Heading
Angle from north in the horizontal plane
Pan
Angle from 'front' in the horizontal plane relative to the camera
Pitch
Angle from the horizontal plane in the camera viewing direction
Roll
Angle from vertical in the camera viewing direction
Field of view
Magnification scaling factor
Range
Distance in metres from the target location
The frame of reference for orientation can either be relative to the world using heading or relative to the camera platform using pan with a specified reference direction such as the front of the vehicle. Pan can be combined with the platform heading to determine camera heading.
Image extent can be determined by either field of view which defines the image scaling factor or range which determines this scaling factor from the camera resolution.
Question: How best can camera attributes be recorded to enable the location of observed objects to be calculated over time?
Background Previous discussions in W3C Spatial Data on the Web (w3c/sdw#1137) and OGC GeoPose (opengeospatial/GeoPose#2) have helped to identify suitable attributes for image characteristics and camera orientation which are aligned with the proposed GeoPose standard.
Use Cases
Dashcam. A forward-facing camera mounted on a vehicle calculates location from a global navigation satellite system (GNSS) and heading from a compass, and these data can be captured in timed video metadata. Orientation only needs to be calculated in 2D as the vehicle is on the ground and low precision is sufficient as the camera has a wide field of view. Multiple cameras can be added to include concurrent rear and side views that may be beneficial to a human driver and which enable an autonomous vehicle to monitor the movement of nearby moving and static objects such as other road users, pedestrians and obstacles to avoid collision.
Drone with gimballed camera. An unmanned aerial vehicle (drone) calculates location from GNSS, height from an altimeter, orientation from a compass and gyro, and camera orientation from the gimbal controller. 3D location and orientation are required as the camera is airborne and more precision is needed due its zoom capability which can affect the field of view.
CCTV. A closed circuit television (CCTV) system includes multiple static cameras which are either fixed or have pan-tilt-zoom (PTZ) capabilities such as roadside traffic cameras. Moving objects can be sequentially tracked by multiple cameras or concurrently observed from multiple angles where fields of view overlap.
Requirements The location and orientation should enable calculation of the camera's viewpoint and allow the location of observed objects to be estimated.
The frame of reference for orientation can either be relative to the world using heading or relative to the camera platform using pan with a specified reference direction such as the front of the vehicle. Pan can be combined with the platform heading to determine camera heading.
Image extent can be determined by either field of view which defines the image scaling factor or range which determines this scaling factor from the camera resolution.