LiDAR Misidentification, How to Handle?
Author： Neuvition, IncRelease time：2023-01-17 02:00:09
LiDAR can provide high-precision point cloud information. With the high-precision point cloud information, the difficulty of many automatic driving problems such as vehicle positioning, high-precision mapping, obstacle detection, and tracking is greatly reduced. However, there is a fatal problem that is LiDAR misidentification happens in the environment such as rain, fog, dust, etc.
1. What is LiDAR misidentification?
In unmanned driving applications, LiDAR is mainly used to detect obstacles on the road. However, due to the inherent characteristics of LiDAR, LiDAR does not work smartly in rain, fog, and dust environments, because rain, fog, and dust will be recognized as obstacles. The root cause of this problem is that the beams of several LiDARs will reflect when they hit the rain, fog, and dust. When the LiDAR receives the reflected laser light, it will judge the rain, fog, and dust as obstacles, resulting in LiDAR misidentification. LiDAR misidentification caused by rain, fog, and dust has become a difficulty in the application of LiDAR.
2. Feasible solutions to handle LiDAR misidentification
At present, to solve the problem of LiDAR misidentification caused by rain, fog, and dust, there are many commonly used methods as follows:
(1) LiDAR using multiple echoes
Multiple echo technology is used in many LiDARs. Multiple echoes refer to the reflection and transmission of the light beam emitted by the LiDAR when it hits some objects, and the transmitted light will also be reflected when it encounters an object. In this way, a beam of light emitted by LiDAR not only receives the reflected light once but also receives multiple reflected lights. This is the multiple echo technology if we use the corresponding algorithm to process multiple emitted lights. The most common processing method is to use the last echo. If in the environment of rain, fog, and dust, the laser can penetrate the rain, fog, and dust, finally hit the obstacle, and reflect it. The last echo can penetrate the rain, fog, and dust, and is reflected when it finally encounters the obstacle; in this way, the last echo can filter out rain, fog, and dust. Of course, the effect of multiple echoes is limited. If there is heavy rain, fog, and dust, the laser cannot penetrate completely, which will still cause LiDAR misidentification.
(2) Using surface-emitting LiDAR
For line-LiDAR, the emitted light is a small beam, which is easily affected by rain, fog, and dust. If there are raindrops or dust particles in the path of the beam, the light beam will be directly reflected, causing LiDAR misidentification. With surface-emitting LiDAR, the probability of being interfered with by rain, fog, and dust will be much lower. This is because the beam becomes thicker. Raindrops and dust particles only emit a small part of the laser. If with the judgment of the intensity of the emitted laser, the interference from raindrops, dust particles, etc. could be filtered out. At present, the LiDAR with FLASH structure adopts the vertical-Cavity Surface-Emitting Laser (VCSEL, or Vertical-Cavity Surface-Emitting Laser), which is a semiconductor whose laser is emitted perpendicular to the top surface. It differs from an edge-fired laser, which emits the laser from the edge.
(3) Using Perceptual Fusion
Using perceptual fusion is also an effective way to solve LiDAR misidentification. It is feasible to use a variety of sensors in the perception system and use the recognition results of other sensors to make up for the defects of LiDAR. For example: using millimeter wave radar. Unlike LiDAR, millimeter wave radar has a strong ability to penetrate rain, fog, and dust, and is hardly affected by rain, fog, and dust. The fusion of the recognition results of the millimeter-wave radar and the recognition results of the LiDAR can also effectively filter out false recognition caused by rain, fog, and dust.
Object detection for transportation systems
In transportation systems, to ensure vehicle and passenger safety and to develop electronic systems that deliver driver assistance, understanding the vehicle and its surrounding environment is essential. LiDAR systems play an important role in the safety of transportation systems. Many electronic systems which add to driver assistance and vehicle safety such as Adaptive Cruise Control (ACC), Emergency Brake Assist, and Anti-lock Braking System (ABS) depend on the detection of a vehicle’s environment to act autonomously or semi-autonomously.
Success use cases in V2X and Railroad projects
In the rail transit fields, we have established in-depth cooperation with the top 10 rail transit integrators and solution providers in China, Europe, and the United States, to provide comprehensive rail transit safety solutions as a trusted partner.
What does the LiDAR point cloud look like in the rail transit application scenarios?
In V2X & smart road scenarios, we have a lot of customers using our LiDAR on their projects to improve traffic management and ensure safety on the road.
Neuvition will continue to work with our customers and partners to develop more and more innovative products and services that solve real-world problems, improve efficiency and productivity, and even save lives.
If you are interested in Neuvition LiDAR solutions or you would like to customize LiDAR solutions for your projects to meet your specific needs, please contact us at firstname.lastname@example.org.