Classes
The following classes are available globally.
-
Object encapsulating work with camera device.
See moreDeclaration
Swift
open class CameraVideoSource : ObservableVideoSource
-
Helper class handling observers: storing, releasing, notifying. Observers are held weakly by the instance of the class. You may inherit your video source from this class to avoid handling observers yourself.
Warning
The implementation uses a non-recursive lock, thus you must not calladd(observer:)
orremove(observer:)
methods fromnotify
closure.Declaration
Swift
open class ObservableVideoSource : NSObject, VideoSource
-
Declaration
Swift
public class BaseVisionManager : VisionManagerProtocol
extension BaseVisionManager: VisionDelegate
-
The main object for registering for events from the SDK, starting and stopping their delivery. It also provides some useful functions for performance configuration and data conversion.
Lifecycle of VisionManager :
create
start
startRecording
(optional)stopRecording
(optional)stop
, then lifecycle may proceed withdestroy
orstart
destroy
Declaration
Swift
public final class VisionManager : BaseVisionManager
-
VisionReplayManager
is a counterpart ofVisionManager
that uses recorded video and telemetry instead of realtime data. Use it to debug and test functions that use Vision in a development environment before testing in a vehicle. Use it in the same workflow as you useVisionManager
after creating it with specific recorded session.Lifecycle of VisionReplayManager :
create
start
stop
, then lifecycle may proceed withdestroy
orstart
destroy
Important
This class is intended for debugging purposes only. Do NOT use session replay in production application.Declaration
Swift
public final class VisionReplayManager : BaseVisionManager
-
Class that encapsulates image buffer and its format.
See more -
Object representing available sign information.
See more -
Single result of sign instance classification.
See more -
Coordinate in Mercator.
This coordinate system is used to locate an object’s geographic location as it would appear on a map. Examples include the GPS position of the ego-vehicle or the position of landmarks, which can be used for localization or position of the detected objects.
Currently we use Spherical Earth Model for geodesic calculations. Each point is specified using
longitude, latitude
.Longitude ranges from
-180
to180
degrees, where0
is Greenwich meridian, the positive direction (+
) is to theEast
, and the negative direction (-
) is to theWest
.Latitude ranges from
See more-90
to+90
degrees, where0
is Equator, the positive direction (+
) is to theNorth
, and the negative direction (-
) is to theSouth
. -
Segmentation result for a single frame.
See more -
Cubic bezier spline in the world space.
See more -
Lane edge description.
See more -
Intrinsic camera parameters representing the source of video frames.
See more -
All detections for a single frame.
See more -
Object aggregating information about road markup and road geometry.
See more -
Description of the object in the world.
See more -
Presents segmentation and detection events emitted from
VisionManager
as well as raw video frames.Displayed content depends on the current value of
visualizationMode
.In
Segmentation
andDetections
modesVisionPresentationViewController
displays according events whenVisionManager
orVisionReplayManager
is started.In
Clear
mode raw video frames are displayed:- in case of
VisionManager
: fromVideoSource
provided duringVisionManager
creation on new events, - in case of
VisionReplayManager
: from recorded video whenVisionReplayManager
is started.
- in case of
-
Object aggregating information about objects and their position in the world around a vehicle. Description includes static and dynamic objects.
See more -
Location of our vehicle with meta information.
See more -
Image object with attributes.
See more -
Single lane object.
See more -
Point in the ISO coordinate system (unit is a meter).
This coordinate system is used to represent the position of an object relative to the device camera’s position in physical space. The origin of this coordinate system is a point projected from a camera position to a road plane.
See more -
Single result of object detection.
See more -
Object representing the state of the camera.
See more -
Represents pixel coordinates on image. Top left corner of an image has (x: 0, y: 0) coordinate.
This coordinate system is used to represent the position of an object relative to the frame. The origin is the left top corner of the frame. The position of an object is expressed as
See morex, y
. -
Object representing geographical coordinates.
See more -
Size of an image.
See more -
All sign classifications for a single frame.
See more