Page 16 - 卫星导航2021年第1-2合期

El‑Sheimy and Li Satell Navig (2021) 2:7 Page 6 of 23

Fig. 1 Part of PLAN sensors on an autonomous vehicle

Table 5 Comparison of traditional map and HD map
Time Accuracy Generation method Challenge
Traditional map (since 2000) Meter‑level to decimeter‑level (Tom‑ Satellite imagery, land‑based mobile Not signifcant
Tom 2020; Liu et al. 2020), road‑level mapping, and onboard GNSS crowd‑
sourcing
HD maps (since 2010) Centimeter‑level (Liu et al. 2020), A team of professional cars equipped High requirements on accuracy, reso‑
where‑in‑lane level with high‑end LiDAR, cameras, GNSS, lution, and updating rate; high data
and INS (Synced 2018) and computational loads
Crowdsourcing using car‑mounted
cameras (Abuelsamid 2017) (chal‑
lenging)

Fig. 2 Comparison of camera, LiDAR, and RADAR performance

make LiDAR a strong candidate in providing high-defni- rotation mechanism on the top of the carrier, which may
tion surrounding environment information. have a problem in its life span. Some manufacturers try to
Te main challenges of using LiDAR are the high price use solid-state LiDAR to alleviate these problems. Apple
and large size. Also, the current LiDAR system has a