georeference digitize

Digitization

Source: NAMRIA

Digitization Procedure

� Display the georeferenced image

� Set the map units to meters

� Create a new theme/layer

� Zoom to the image until pixels are big enough to be seen

� Set the appropriate snap tolerance to automatically connect two lines

Digitization Procedure

� Start digitizing. Digitize in the middle or center of the feature. A tolerance limit of equal to and less than

0.25 mm (equal to 12.5m error) away from the midpoint is acceptable.

� Make an attribute table corresponding to the features created.

� Edit feature, if necessary.

4.235 m. (0.0847 mm)

Digitization

4.235 m.

(ground)

4.235 m.

(ground)

4.235 m.

(ground)

Projections, Georeferencing and Geometric Rectification

Source: NAMRIA

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Digitization

Georeferencing/Geometric

Rectification� Georeferencing is establishing relationship

between an image coordinate system (row, column) and a map coordinate system (x, y or longitude, latitude or Easting, Northing). It is also the process of assigning map coordinates to image data.

� Geometric rectification is the process by which image or grid is converted from image coordinates to real-world coordinates.Ittypically involves rotation and scaling of grid cells, and thus requires resampling of values.

Georeferencing/Geometric

Rectification� The pixels of the new grid will probably not align

perfectly with the pixels of the original grid, so

resampling is needed.

� Resampling is the process of extrapolating data

values for pixels on the new grid from the values

of the source pixels.

� Rectification, by definition, involves

georeferencing, since all map projection systems

are associated with map coordinates.

Procedure

� Rectification is accomplished by

– Matching points of known location in both image and map or image coordinate system

– These points are used to compute the best fit polynomial transformation.

At least, the 4 corners should be

selected as control points for georeferencing.

10’ x 15’ Map

In selecting more points in additionto the 4 corner ticks, make sure that thereference pointsare well-distributed.

Selecting / Locating Points

� Intersection of roads and other man-made features that can to located in both map and image.

� GPS points of ground locations that can be identified in the image.

� Points should be located evenly across the image (sometimes difficult in natural areas)

Transformation Models

Transformation Models Minimum No. of Points

Affine Transformation 4 points

Second Degree Polynomial 6 points

Affine Transformation

Affine Transformation

� A linear correction model that involves three parameters for X and Y.

� The three parameter represent shifts in

– X or Y direction

– Scale

– Rotation about a point.

Root Mean Square� RMS is the distance between the input (source)

location of a control point and the retransformed location for the same control point.

� The smaller the RMS the better. It is the measure of best fit to the actual locations of map coordinates.

� RMS error is normally expressed as a distance in pixel width.

– If 1 pixel for a digital 1:50,000 topographic map scanned at 300 dpi is equivalent to 4.325 m. on the ground, then an RMS error of 2 will be equal to 2 x 4.325 m = 8.65 m. on the ground.

Root Mean Square

Root Mean Square

Root Mean Square

Root Mean Square

Root Mean Square