Title :
Segmentation of Viscose Filament Fracture Surface Image Based on SOFM Network Fusion
Author :
Zhang, Jiaquan ; Feng, Yi ; Lin, Xiaolong ; Zhang, Tieqiang ; Zhang, Weiyi
Author_Institution :
Jilin Univ., Changchun, China
Abstract :
A method of viscose filament fracture surface image segmentation based on SOFM network fusion is proposed. Firstly, the binarization method based on two-way weighting sequential smooth is used to segment the two images which are obtained in different light intensity. Secondly, fuse the two binary images based on SOFM fusion. After that, the final segmentation image is obtained. Using this method can record the information of viscose filament fracture surface accurately. The experimental results show that this method is much better when we can´t get clear fiber images. At the same time, it provides a way to segment the ununiformity brightness images.
Keywords :
fracture; image fusion; image segmentation; optical fibres; self-organising feature maps; viscosity; SOFM network fusion; SOFM neural network; binarization method; light intensity; two-way weighting sequential smooth; ununiformity brightness images; viscose filament fracture surface image segmentation; Artificial neural networks; Entropy; Fuses; Image segmentation; Optical fiber networks; Pixel; Surface cracks;
Conference_Titel :
Information Engineering and Computer Science (ICIECS), 2010 2nd International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-7939-9
Electronic_ISBN :
2156-7379
DOI :
10.1109/ICIECS.2010.5678395
