Luminar Investor Presentation Deck
DEEP DIVE: PERCEPTION SOFTWARE
To deliver highway automation and proactive safety, a vehicle needs to:
Understand the complete scene,
in any environment
SEMANTIC SEGMENTATION
Assigns a "type" to each object in a scene:
examples include driveable road, non-
driveable space, lanes, etc
Luminar value add: the unique ability to
understand a complete scene in 3D in one
single sensor
The
Detect and classify objects at
long distance, in all lighting and
weather conditions
OBJECT DETECTION & CLASSIFICATION
Detect objects in a scene, at a long
distance, and classify those objects as
vehicle, pedestrians, and cyclists
Luminar value add: provides more time at
high highway speeds for planning and
decision making at night and in inclement
weather/environments
ph
Track lane and road markings,
identifying driveable and non-
driveable space
|||
LANE & ROAD TRACKING
Detects and classifies lane markings and
road edges; classifies single, double, dotted
and dashed lines; assigns objects a lane ID,
even in the absence of lane markings
Luminar value add: provides information for
path planning and decision making; extends
road and lane capture beyond the ability of
headlights, a failsafe operation if headlights
are not functioning properly
26View entire presentation