Title : 
Robust Epiphyseal Extraction Method Based on Horizontal Profile Analysis of Finger Images
         
        
            Author : 
Park, Ki-Hong ; Lee, Jong-Min ; Kim, Whoi-Yul
         
        
            Author_Institution : 
Hanyang Univ., Seoul
         
        
        
        
        
        
            Abstract : 
By existing epiphyseal extraction methods, epiphyses are not extracted accurately, because an error occurs in the detection of the finger´s central axis. This paper proposes a new, more accurate epiphyseal extraction method that estimates the central axis from the top of the distal phalange to the upper part of the metacarpal. The method entails analysis of a finger´s horizontal profile that detects that finger´s central axis, on and from which epiphyses are detected and extracted, respectively. The result of an experiment showed that this method accurately extracts epiphyses even from finger images for which the detection of the central axis is difficult.
         
        
            Keywords : 
bone; estimation theory; feature extraction; medical image processing; object detection; central axis estimation; finger central axis detection; finger images; horizontal profile analysis; robust epiphyseal extraction; Biological tissues; Bones; Data mining; Failure analysis; Fingers; Image analysis; Image edge detection; Information analysis; Information technology; Robustness;
         
        
        
        
            Conference_Titel : 
Information Technology Convergence, 2007. ISITC 2007. International Symposium on
         
        
            Conference_Location : 
Joenju
         
        
            Print_ISBN : 
0-7695-3045-1
         
        
            Electronic_ISBN : 
978-0-7695-3045-1
         
        
        
            DOI : 
10.1109/ISITC.2007.25