This paper presents an algorithm for the acquisition of height value of arbitrary points on contour map. Acquired map image is processed by a thinning process. Through a local feature classifier of legend, its position and digit string are extracted. ...
This paper presents an algorithm for the acquisition of height value of arbitrary points on contour map. Acquired map image is processed by a thinning process. Through a local feature classifier of legend, its position and digit string are extracted. Shorter lines are generated by removal of neighboring pixels of diverging point and lines shorter than threshold are eliminated as noise. To connect the disconnected contour line, end point is connected to another one nearest to adjacent end point centered by summit and not cross the another contour line. Decreasing height value with equal ratio is assigned to every contour line met by the left and right direction scan line from summit. To calculate height value at a point, distance ratio between it and neighboring contour lines or between summit and contour line is considered.
Employing these steps, proposed algorithm successfully assigns height value to arbitrary points. Acquired information can be applied to the generation of detailed contour map and 3D reconstruction of it.