Reprojection, Georeferencing, Control Points and Transformation

Question 1

Georeferencing

Answer 1

Georeferencing is to establish a relationship between page coordinates on a planar map and known real-world coordinates.

Or other way is the act of assigning geographic locations to features of the spatial data that do not have any real world coordinates.

Question 2

Geometric Transformation

Answer 2

The process of converting a digitalized map, satellite image or aerial photograph from one coordinate system to another by using a set of control points and a transformation equation.

Georeferencing and transformation are often subsequent process of making data useable in GIS

GIS software package usually provide these methods.

Affine transformation method is most commonly use..

Question 3

Transformation involves

Answer 3

Scaling, rotation, translation, and skew a dataset to a given set of geographic or projected coordinates.

Step 1: Updates the control points to real-world coordinates.

Step 2: Uses the control points to run a transformation.

Step 3: Creates the output by applying the transformation equations to the input features.

Question 4

Good Control Points

Answer 4

Select the features clearly and neatly visible on the source map/satellite image.
Road Intersections and Pointed landmarks.

The real-world coordinates are known:

May be available from existing source.

Acquire from an authentic map or GPS survey.

Question 5

Transformation Methods

Answer 5

Many mathematical modes (equations)

Each method distinguished by the geometric properties it can preserve and the changes it allows.

The changes could be: change of position and directly, change of scale, and changes in shape and size.

Commonly used:

1. Equiarea
   - Transformation allows rotation and preserves shape and size.
2. Similarity
   - Transformation allows rotation and preserves shape but not size.
3. Affine
   - Transformation allows angular distortion but preserves parallelism.
4. Projective
   - Transformation allows both angular and length distortion. So a rectangle to be transformed into an irregular quadrilateral.

Question 6

The Process of Transforming All Of Your Datums TO Match One Measurement System Is Called:

Answer 6

Reprojection.

Question 7

Georeferencing Is The Process Of Aligning An Unreferenced Data Set With One That Has Spatial References Information

Answer 7

True

Question 8

The Common Areas That Tie Unreferenced Data to Spatially Referenced Data Are Called

Answer 8

Control Points

Question 9

Resampling of Pixel Values

Answer 9

Result of geometric transformation of a image is a new image based on a given coordinate system.

New image has no pixel values. These must be filled through resampling.

Resampling refers to filling of each pixel of new image with a value derived from original image.

Various methods exist.

Question 10

Three Common Resampling Methods

Answer 10

1. Nearest neighbor resampling: fills each pixel of the new image with the nearest pixel value from the original image.
2. Bilinear interpolation method: Uses the average of the four nearest pixel values from the three linear interpolations.
3. Cubic Convolution: Uses the average of the 16 nearest pixel values from five cubic polynomial interpolations.

Question 11

Affine Transformation

Answer 11

The equations used in Affine Transmissions are:
X= Ax + By + C
Y = Dx + Ey + F

X, Y are input coordinates and X, Y are output coordinates.

Coefficient C represents translation in the X direction, and coefficient F the translation in the Y direction.

Coefficients A, B, S, and E are related to rotation, skew, and scaling.

Allows rotation, translation, skew, differential scaling while preserving line parallelism.

Question 12

Affine Transformation Properties

Answer 12

The equations requires at least three known points to estimate its six coefficients.

The known points are also knows as tics/ground control points (GCPs)

At least four known points are commonly used for reducing problems with measurement errors and to allow for a least-squares solution.

From the Root Mean Square (RMS) error value is the indicator for the goodness of control points that derives from the least-square equation.

Question 13

Polynomial Order In Transformations

Answer 13

1. 1st Order Polynomial.
2. 2nd order polynomial.
3. 3rd Order Polynomial.

Question 14

1st Order Polynomial

Answer 14

Requires a minimum of 3 displacement links, but should have more even though 3 gives RMSE=0!.

Is a homogenous transformation: only shifts origin, scales and rotates.

Straight lines will be preserved.

Question 15

2nd Order Polynomial

Answer 15

Requires 6 points (displacement links) minimum.

Is a differential transformation so its “warps” the raster.

Straightlines on raster may no longer be straight.

Question 16

3rd Order Polynomial

Answer 16

Requires 10 points minimum.

Question 17

All of the following are poor control-point selection choices except:

• center of a corn field
• intersection of narrow road
• a point o the shoreline of a beach.
• a corner of a tall building seen on an image.

Answer 17

Intersection of a narrow road.

Question 18

When selecting control points, it is best to cluster them in one area of the map. True or False.

Answer 18

False

Question 19

In georeferencing, what is the minimum number of control points required to fit an unreferenced image to the source?

Answer 19

3

Question 20

Root Mean Square (RMS) error

Answer 20

The root mean square (RMS) error is a common measure of the goodness of control points.

It measures the average deviation between the actual (true) and estimated (digitalized or selected) locations of control points.

The RMS error is derived for the equation.

Question 21

Misinterpretation of RMS Error on Digitized Maps

Answer 21

The thin lines represent correct soil lines and the thick lines incorrect soil lines.

[ The x values of the upper two tics were decreased by 0.2’’ while the x values of the lower two tics were increased by 0.2’‘ ]