4.2 Geometric correction

Question 1

Geometric correction

Answer 1

• The Earth has a spherical shape. Through map projection, spherical Earth’s surface is transformed into maps.
• The remote sensor images the Earth s surface without knowledge of map projections. It merely records DNs and the location of each pixel is not measured by the sensor.
• Geometric correction accounts for distortions in the image.
• Georeferencing the image relates the pixel position in the image to its corresponding position on the ground.
• Knowing the relation between pixel position and ground position, we can transform any pixel position to the ground position.

Question 2

Georeferencing

Answer 2

• Georeferencing solves two problems:
– Map coordinates of features identified in the images.
– The geometric distortions of the RS image are corrected.
• Geometric correction accounts for distortions in image due to motions of platform and scanner mechanism or curvature of the Earth.
• Geometric distortions are more apparent in large FOVs.

Question 3

Geometric distortions

Answer 3

•Internal distortions
–Depend on the sensor or platform

•External distortions
–Independent of sensor and platform
–Usually vary in space and time

•Systematic distortions
–Predictable
–Pre defined and mathematically modelled

• Non-systematic distortions
–More difficult to correct
–Most important source: Random movement of platform and Relief displacement
–Digital Terrain Model (DTM) and Ground Control Points (GCPs) are used in empirical models to correct Non systematic distortions.

•Earth rotation effect
–Earth rotation affects space borne scanner and line camera images thathave a large spatial coverage.

Question 4

Image distortions

Answer 4

• Scan skew distortion
– In Whiskbroom sensors
– While the scan mirror completes one active scan, the satellite moves along the ground track.
– Scanning is not at right angle to the satellite track but is slightly skewed.
– Produces across track geometric distortion

Question 5

Cause of Geometric distortions

Answer 5

•Oblique viewing
–If the camera is tilted, the image scale is non uniform.
–Objects in the foreground appear larger than those farther away from the nadir.

•Relief displacement
–Caused when there is significant terrain relief.
–Terrain points whose elevation are are above or below the reference elevation are displaced in the image.
–Causes buildings or trees to lean away from the nadir point.
–Results in inaccurate or wrong map coordinates.

•Altitude change
–Change in altitude causes change in GIFOV
–Rotation around three axes
§Roll: rotation about the axis running from nose to tail
§Pitch: nose up or down about the axis running from wing to wing
§Yaw: nose left or right about the axis running up and down

*note: Nadir point: the point on the ground directly below the satellite

Question 6

Orthorectification

Answer 6

The process of removing internal and external distortions including
–Satellite/aircraft motion
–Terrain relation geometric distortions (relief displacement)

Question 7

Orthorectification methods

Orthorectification methods mean we convert image by image that looks like every pixel at the nadir point

Answer 7

– Rigorous model (camera model)
• Physical based model
• Takes into account a wide variety of factors influencing the image acquisition

– Rational Polynomial Coefficients (RPCs)
• Does not require knowledge of all factors influencing the image
• Easy to implement

Distances, areas, and directions are more accurate in the orthorectified image.