Photogrammetry, Remote Sensing, and Mapping
What this domain tests
Whether you can specify an aerial or lidar mapping product and check that it meets a stated accuracy, and whether you understand what an image or point cloud can and cannot deliver.
Photogrammetric relations
- Photo scale = f / H, where f is focal length and H is flying height above ground.
- Relief displacement d = r × h / H, where r is the radial distance from the principal point and h is object height.
- Standard aerial photography uses about 60% endlap and 25–30% sidelap to support aerotriangulation and stereo viewing.
- Ground sample distance sets the finest detail; it is not the same as map accuracy.
Lidar and remote sensing
Lidar delivers a classified point cloud; the bare-earth class becomes a DTM, and the first-return surface is a DSM. Contours belong on a DTM. Multispectral imagery supports indices such as NDVI. Every sensor has spatial, spectral, radiometric, and temporal resolution.
Mapping accuracy
ASPRS-style accuracy is stated as RMSE in z and horizontal RMSE at 95% confidence. A common check is that contour interval should be roughly two to three times the vertical RMSE, not equal to it.
Worked example
A camera with f = 6 in (0.50 ft) flies at 6,000 ft above ground: scale = 0.50 / 6,000 = 1:12,000, so 1 in on the photo equals 1,000 ft on the ground. A tree 100 ft tall standing 3.00 in from the principal point is displaced d = 3.00 × 100 / 6,000 = 0.05 in, which at 1:12,000 is about 50 ft on the ground - enough to misplace a fence line if you digitize from a single photo.
Traps that cost marks
- Using elevation above sea level for H instead of height above ground.
- Assuming a DSM will produce clean bare-earth contours.
- Treating GSD as a guarantee of horizontal accuracy.
- Digitizing boundary evidence from uncorrected imagery instead of a controlled orthophoto.