What Is LiDAR?
Benefits of LiDAR
LiDAR Uses
There are multiple uses for LiDAR.
- In atmospheric science, it can improve weather forecasting through cloud detection
- Topographic LiDAR can map buildings in 3D, e.g., archaeological sites
- Green-light LiDAR can map the ocean floor
- In automotive applications, LiDAR provides a very high-resolution measurement of object distance, improving ADAS and facilitating autonomous driving
How LiDAR Works
LiDAR emits a laser light pulse towards objects, typically in the near-infrared region. The light reflected is detected by a receiver. The time taken for the reflection to be received can is used to calculate the distance to the object. A LiDAR emitter can transmit millions of pulses every second.
A LiDAR system can also transmit light in multiple directions using a moving mirror or multiple beams. This enables it to generate a distance map, showing how far all objects in a particular direction are from the vehicle. This map can be used to construct a 3D model of the surroundings, with sensor fusion employed to improve the model’s accuracy by adding data from additional sources. By tracking object distances over time, LiDAR can reveal whether the objects are approaching or moving away, making it possible to predict an impending collision. This can enable ADAS to take timely avoidance actions.
LiDAR vs. Radar and Cameras
While LiDAR emits and detects laser light, radar operates similarly using a different range of the electromagnetic spectrum—radio waves. Due to the wavelengths used, LiDAR offers higher resolution and better accuracy than radar. While radar can also operate in inclement conditions, it is more susceptible to interference than LiDAR and has a lower resolution. On the other hand, radar can penetrate objects with low density and is cheaper to implement than LiDAR. It might, for example, be able to detect solid objects beyond smoke, which could foil LiDAR.
Cameras, in contrast, do not detect signals emitted from the vehicle but collect environmental light reflected from objects. The result is a 2D image without direct 3D distance modeling. However, camera images can include information other than distance that can be used to generate a 3D model of the world, particularly if the images from multiple cameras located at different places on the car are combined. With sophisticated algorithms, the inputs from radar, LiDAR and cameras can combine to generate a comprehensive model for advanced autonomous driving applications.
Check Out Our Other Ultimate Guides
