Title :
Melanocytic lesion images segmentation enforcing by spatial relations based declarative knowledge
Author :
Kwasnicka, Halina ; Paradowski, Mariusz
Author_Institution :
Inst. of Appl. Informatics, Wroclaw Univ. of Technol., Poland
Abstract :
This paper presents a method for enforcing image segmentation. Method presented here is a part of a wider research - an image understanding system. Regions of the image, extracted by a basic segmentation process, belong to certain classes. Spatial relations are calculated for extracted regions of the image, according to the declared knowledge. Declarative knowledge in the form of semantic rules is used as a second step to eliminate errors in the basic segmentation process. The approach is generic, but in this case it is used for melanoma skin lesions diagnosis.
Keywords :
cancer; feature extraction; image segmentation; knowledge based systems; medical image processing; patient diagnosis; skin; image region extraction; image understanding system; melanocytic lesions image segmentation; melanoma skin lesions diagnosis; semantic rule; spatial relation based declarative knowledge; Artificial intelligence; Histograms; Image segmentation; Informatics; Lesions; Logic; Malignant tumors; Neural networks; Production; Skin cancer;
Conference_Titel :
Intelligent Systems Design and Applications, 2005. ISDA '05. Proceedings. 5th International Conference on
Print_ISBN :
0-7695-2286-6
DOI :
10.1109/ISDA.2005.63