Full text: XIXth congress (Part B3,1)

  
Carsten Garnica 
  
A NEW APPROACH TO EDGE-PRESERVING SMOOTHING FOR EDGE EXTRACTION AND 
IMAGE SEGMENTATION 
Carsten GARNICA, Frank BOOCHS, Marek TWARDOCHLIB 
i3mainz, Institute for Spatial Information and Surveying Technology 
FH Mainz, University of Applied Sciences 
Holzstrasse 36, D-55116 Mainz, Germany 
carsten (? eeoinform.fh-mainz.de 
http://www.i3mainz.fh-mainz.de 
  
WG III/3 : Feature Extraction and Grouping 
KEY WORDS: Image Processing, Filtering, Radiometric Quality, Feature Extraction, Edge Extraction, Segmentation, 
Multi-spectral Data. 
ABSTRACT 
This paper introduces a new approach to edge-preserving smoothing of digital images. It is designed for an effective 
elimination of the image noise within digital images what is an essential step in applications like edge detection or 
image segmentation. 
The approach presented here tries to overcome some of the disadvantages of existing smoothing filters and is conceived 
as an extension of the edge-preserving Maximum Homogeneity Neighbour Filter. The algorithm cleans up the image 
noise in the homogeneous areas, but preserves all image structures like edges or corners. It is shown that the new filter 
algorithm combines the advantageous features of different types of filters. The algorithm is not only applicable to 
grayscale images, but can be extended to multi-channel data, like color images too. 
The performance of the algorithm is achieved by a more complex and differentiating treatment of the image data 
compared to conventional concepts. 
KURZFASSUNG 
Es wird ein neuer Ansatz zur kantenbewahrenden Gláttung von Digitalbildern vorgestellt. Wenn Merkmale wie z.B. 
Kanten aus einem Digitalbild extrahiert werden sollen oder eine Bildsegmentierung durchgeführt werden soll, muss 
zuerst ein Filterprozess angewendet werden, um das Bild vom Rauschen zu sáubern. 
Der hier vorgestellte Ansatz versucht, einige der Nachteile der bestehenden Gláttungsfilter zu überwinden, indem der 
Maximale Homogene Nachbar Filter erweitert wird. Der Algorithmus beseitigt das Bildrauschen in den homogenen 
Bereichen, wáhrend alle Bildstrukturen wie Kanten oder Ecken bewahrt bleiben. Es wird gezeigt, dass der neue 
Filteralgorithmus die erwünschten Eigenschaften verschiedener Filtertypen vereinigt. Der Algorithmus kann nicht nur 
auf Grauwertbilder angewendet werden, sondern auch auf Mehrkanaldaten wie z.B. Farbbilder. 
Die Leistungsfáhigkeit des Algorithmus basiert auf einer komplexeren und differenzierten Behandlung der Bilddaten als 
bei bisherigen Konzepten. 
1 INTRODUCTION 
1.1 Motivation 
This paper introduces a new approach to edge-preserving smoothing of digital images as a preprocessing step for 
feature extraction and/or image segmentation. Although there are many filtering algorithms available, the results of 
these algorithms are not satisfying for all applications. In specific cases of object detection, as to be performed in an 
ongoing research project, standard algorithms may have deficiencies. Here, the detection process has to work even 
under bad conditions, making it necessary to have a reliable algorithm. The objects to be detected within the digital 
images are sometimes very small (eg. down to 3*3 pixels), the contrast is sometimes low and considerable radiometric 
noise has to be allowed. 
  
320 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 
  
  
F^ Fn rm OY rr P? UM eed 
5. 
b à 
— 
"m m 0. 71]
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.