Background subtraction is a widely used approach for detecting moving objects in point clouds captured by static LiDARs. For real-time detection and recognition it is crucial to reduce the amount of data to be processed and to be able to segment different objects. This is achieved by removing the background of a point cloud and extracting the foreground.
Beam deflection unit
The beam deflection unit is a core component of scanning LiDARs. It directs the laser pulses over the entire field of view. This is done using e.g. mechanical or micro-electromechanical system technologies. Blickfeld uses a silicon-based (MEMS) beam deflection unit. (link https://www.blickfeld.com/blog/mems-technology-for-lidar/)
Coaxial design (monostatic design)
A coaxial design means that the axes of the laser beam and the receiver optics coincide. Thus, emitted laser rays are returned back to the detector on identical paths. Thanks to the coaxial setup, daylight, background light or beams of other sensors are suppressed and a high signal-to-noise ratio is achieved.
Collimation describes light with low divergence. Perfectly collimated light would mean parallel beams and plane wavefronts. Due to diffraction this is not possible in reality. In LiDAR systems usually the size of the emitter and the limited available optical aperture limit collimation.
The convex hull describes the smallest possible circumference of a body by connecting the outermost points of the object.
If multiple laser pulses (e.g from multiple LiDAR sensors) are emitted at the same time, the detector may not be able to distinguish between signals, resulting in erroneous range measurements and 3D points. To mitigate cross-talk effects, Blickfeld makes use of spectral filtering and spatial filtering.
Field of view
LiDAR / 3D LiDAR
Silicon photomultipliers (SiPMs)
The Blickfeld setup is coaxial. As a result, only the photons from the exact direction in which the laser was sent are collected. This spatial filtering supresses straylight, achieving a high signal-to-noise ratio.