The image data of multispectral (MS) of 4m resolution and panchromatic (PAN) of 1m resolution can be obtained from IKONOS. The fused MS images of 1m resolution can be obtained by applying the fusion process to the MS images and PAN image. When an existing image fusion process is applied to the IKONOS image, the distortion of the spectrum is appeared. This is because of not being reflected in 1m MS images to obtain the reflection of spectral response between the band of MS and the band of PAN by fusion process. To solve this problem, we proposed an image fusion algorithm using steepest decent method in this paper. Our method has been succeeded to improve correlation coefficients between the original 4m MS image and the fused MS image which present degree of spectral distortion.

View full abstract

In this paper, we propose a method for automatic extraction of lung area and its classification for abnormal area which include ground-glass opacity from thorax CT image sets. In the extraction of lung area, we segment the lung area of the region of interest employing binarization and labeling process from the inputted each slices images. The region having the largest area is regarded as the tentative lung regions. Furthermore, the ground-glass opacity is classified by correlation distribution between successive slices from the extracted lung region with respect to the thorax CT images. Experiment is performed employing twenty six thorax CT image sets and satisfactory recognition rates were achieved.

View full abstract