In: Wagner W., Szekely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
531 
For the obtaining of good results through the use of morphologic 
tools is necessary to choose the structuring element appropriately 
to be used in the functions. The element estruturante is the key for 
the success of the operations, since it is chosen in an appropriate 
way. 
The choice of the element estruturante depends on the geometric 
form object to be extracted in the image. For instance, linear 
structuring elements are appropriate for the linear objects 
extraction. It can also be considered several rotations of the 
structuring element to extract the wanted forms from the objects 
contained in the image. For instance, it is wanted to extract the 
highways, the ideal would be to use a structuring element with 
mask in form of line segment, however this element should be 
rotated in several directions to answer the multiple orientations of 
the highways. The size of the structuring element depends on the 
characteristics that you want to extract in the image. Structuring 
elements larger preserve larger characteristics while smaller 
elements preserve smaller details in the image. Some examples of 
structuring elements are shown in the Figure 1. 
Figure 1. Structuring elements: (a) cross, (b) square and (c) line. 
The type and nature of detected information depend of the 
structuring element and studied image choice. The shape 
and size of the structuring element must be adapted to the 
geometric properties of the image objects to be processed. 
The Mathematical Morphology presents as principle two basic 
operations: Erosion and Dilation, from which all other 
morphological operations are derived. 
a) Erosion: According to Soille (1999), the erosion of a set X by 
a structuring element B is denoted by e «^and is defined as 
the locus of points, x, such that B is included in X when its 
origin is placed at x: 
b) Dilation: According to Soille (1999), the dilation of a set X 
by a structuring element B is denoted by s b (X) and is defined 
as the locus of points x such that B hits X when its origin 
coincides with x: 
S B (*) = {*| B x nX*0} (2) 
For this definition, when the structuring element is verifying 
the image, the neighborhood of the central point should have a 
intersection with the relevant points of the image, capturing 
more pixels. The application of this operator produces the 
following effects: increases the objects in the image, fills out 
small holes and connects close objects. 
Erosion and dilation can be used in a variety of ways, in 
parallel and series, to give other transformations including 
thickening, thinning, pruning and many others. 
c) Thinning: According to Soille (1999), the thinning of a set X 
by a composite structuring element B is denoted by X O B and 
defined as the set difference between X and the hit-or-miss 
transform of X by B (HMT B (X): 
XOB = X\HMT B (X) (3) 
The hit-or-miss transformation of a set X by a composite 
structuring element B=(B], B 2 ), is the set of points, x, such 
that when the origin of B coincides with x, Bj fits X and B 2 
fits X c . 
HMT„(X) = e m (X)ne, 2 (X‘) (4) 
d) Pruning: According to Soille (1999), the pruning 
transformation is implemented through the detection of final 
points and their removal up to the idempotency, in other 
words: 
PRUNE{X) = {XAFIs 
(5) 
£ B (X) = {x\B x cX} (1) 
This definition indicates that the structuring element B slides 
on the image and it compares the neighborhood of each pixel 
with the neighborhood of the central point (that most of the 
time will correspond to the physical center of the structuring 
element) preserving the pixels where the neighborhoods 
coincide. In general, the erosion presents the following effects 
in their results: decrease of particles, elimination of the 
inferior size grains than of the structuring element, increase of 
the holes and it allows the separation of close grains. 
Where E is related to the structuring element used to detect 
final points and °° indicates that the sequential thinning is 
iterated up to stability. 
When the pruning is taken into stability, the only parts of the 
skeleton not suppressed are the closed arcs or “ties” of the 
skeleton. 
2.2. Data Used 
The test image chosen corresponds is a panchromatic image of the 
high resolution satellite QuickBird, with spatial resolution of 61 
cm. This contains as main feature a part of the Raposo Tavares 
highway, located in the area of Presidente Prudente city (SP), 
Brazil. The Figure 2 presents the original image.