Fundamentals of High‑Precision Mapping and Coordinate Systems

High‑precision maps differ from ordinary electronic maps in that they require sub‑meter to centimeter‑level accuracy and contain far richer road‑traffic elements. This brings many challenges in data production that demand professional surveying knowledge.

The article introduces the basic concepts needed to understand high‑precision mapping, starting from the definition of the Earth and moving to various coordinate systems used in the field.

1. How to Define the Earth

1.1 Earth Ellipsoid – An idealized, low‑flattening ellipsoid generated by rotating a short‑axis ellipse around the Earth's short axis.

1.2 Geoid – The physical sea‑level surface extended under the land, which is irregular due to uneven gravity distribution and cannot serve as a precise mathematical model.

1.3 Reference Ellipsoid – A mathematical model that best fits the geoid, defined by six parameters (latitude, longitude, height, two axis‑offset components, and azimuth).

1.4 Geocentric / Geodetic Coordinate Systems

Geocentric (ECEF): XYZ axes with origin at the Earth’s centre of mass.

Geodetic (BLH): Latitude, longitude, height referenced to a chosen ellipsoid.

1.5 Common Datums

WGS‑84: International standard ellipsoid (a = 6378137 m, b = 6356752.314 m, 1/f = 298.257223563).

Beijing 54, Xian 80, GCJ‑2000: Regional datums used in China.

2. Defining a 2‑D Map

2.1 Map Projection – Mathematical methods that transform the curved Earth surface onto a plane, inevitably introducing distortion. Common projections include cylindrical, conic, azimuthal, and others.

2.2 Typical Projections

Mercator / Web Mercator – Conformal cylindrical projection; widely used in web maps.

Gauss / UTM – Conformal transverse‑cylindrical projection adopted by many countries.

3. Coordinate Systems in High‑Precision Mapping

3.1 ECEF (Earth‑Centered Earth‑Fixed) – Origin at the Earth’s centre; Z axis points to the North Pole.

3.2 ENU (East‑North‑Up) Local Tangent Frame – Origin at the vehicle; X points east, Y points north, Z points up.

3.3 Vehicle Body Frame – Origin at the vehicle’s centre of mass; axes follow the right‑hand rule (right‑forward‑up).

3.4 LiDAR Frame – Origin at the LiDAR’s rotation centre; Z axis points upward.

3.5 IMU Frame – Origin at the inertial measurement unit; axes aligned with gyroscope/accelerometer axes.

3.6 Camera Frame – Origin at the optical centre; X right, Y down, Z forward.

3.7 Right‑Handed Coordinate System – Defined by the thumb, index, and middle finger of the right hand.

4. Coordinate Transformations

4.1 Geodetic ↔ ECEF – Conversion between latitude/longitude/height and Cartesian XYZ using the same reference ellipsoid.

4.2 ECEF ↔ Local (ENU) – Translation and rotation to a vehicle‑centred frame.

4.3 Geodetic ↔ Projected (e.g., Mercator, UTM) – Uses map‑projection formulas.

Transformation between different datums often employs the 7‑parameter Helmert (Bursa‑Wolf) model: three translations (ΔX, ΔY, ΔZ), three rotations (Δα, Δβ, Δγ), and a scale factor (m).

5. 3‑D Rotation Concepts

5.1 Euler Angles – Sequential rotations around global Z, then local X, then local Z axes.

5.2 Gimbal Lock – Loss of a degree of freedom when two rotation axes align.

5.3 Rotation Matrix – Matrix representation of arbitrary rotations.

5.4 Rotation Vector (Rodrigues) – Axis‑angle representation.

5.5 Quaternion – Four‑component representation that avoids gimbal lock and is computationally efficient.

6. Supporting Standards and Tools

6.1 EPSG – European Petroleum Survey Group maintains a database of coordinate reference system definitions (SRID).

6.2 WGS 84 / Web Mercator – EPSG:3857 is the de‑facto standard for web maps.

6.3 Proj.4 – Open‑source library for coordinate transformations, available in many languages.

7. References

Links to articles on autonomous‑driving coordinate systems, GIS projection theory, Euler angles, quaternion visualisation, and more.