Translator Disclaimer
26 October 1999 Simultaneous registration and segmentation of images in wavelet domain
Author Affiliations +
A novel method for simultaneous registration and segmentation is developed. The method is designed to register two similar images while a region with significant difference is adaptively segmented. This is achieved by minimization of a non-linear functional that models the statistical properties of the subtraction of the two images. Minimization is performed in the wavelet domain by a coarse- to-fine approach to yield a mapping that yields the registration and the boundary that yields the segmentation. The new method was applied to the registration of the left and the right lung regions in chest radiographs for extraction of lung nodules while the normal anatomic structures such as ribs are removed. A preliminary result shows that our method is very effective in reducing the number of false detections obtained with our computer-aided diagnosis scheme for detection of lung nodules in chest radiographs.
© (1999) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Hiroyuki Yoshida "Simultaneous registration and segmentation of images in wavelet domain", Proc. SPIE 3813, Wavelet Applications in Signal and Image Processing VII, (26 October 1999);

Back to Top