Perception Blog: Technology Comparison -Flash and Scanning LiDAR
A blog post by
David Brodie BSc (Eng), Sr. Project Manager of Perception and AI, LeddarTech®
My name is David Brodie. With this occasional blog, I hope to create a platform to exchange ideas on technology and solution related to perception. I hope to discuss a selection of topics, some at a high level and some in more technical detail.
To start with let’s consider a high-level analysis of one of the classic problems which every autonomous vehicle faces: detecting debris or small objects.
Consider first a classic scanning lidar that surveys its environment using several scan lines. Consider the following figure.
Figure 1: a scanning lidar’s view of the world
Theoretically, all four small objects could be detected. Objects 1 and 3 are detected as they are directly in the path of some scan line. Although object 2 is nearer than object 3 it is not detected as it falls between scan lines. Similarly, object 4 is not detected through in range as it too falls between scan lines. As a scanning lidar approaches a small object, the object will be detected and lost with increasing frequency until it is close enough to be hit by a scan line in every frame.
The following figure shows the same scene surveyed by a flash lidar like LeddarTech’s Pixell.
Figure 2: flash lidars view of the world
In this case, the entire field of view of the lidar is illuminated. Rather than detecting point reflections, the flash lidar detects reflections from a segment. Once an object becomes detectable it is continuously detected as there are no gaps in the illumination. There are challenges here too. Object 4 only fills a small section of the relevant area or “segment”. Object 3 is larger but split across multiple segments and so also only fills a small portion of each segment. It may not be easy to detect but with good beam steering and signal processing, it is possible. Once an object becomes detectable; it is reliably detected from then on.
Return to Latest News and Media Coverage
This website places cookies on your device to give you the best possible user experience. By using our websites, you agree to our cookies being saved on your device (unless you have disabled cookies in your settings). For more information, please read our Terms of Use and our Privacy policy.