The Importance of Features and Primitives for Multi-dimensional/Multi-channel Image Processing
In the context of image processing, a major role is played by the features and primitives that describe the data under examination and on which the processing operation is performed. Images acquired by different sensors, for different parameter values tunings, and multi-dimensional and multi-temporal data are becoming easily available, thus increasing the dimensionality of the classification space, then the need for feature-selection techniques.
KeywordsArtifical Neural Network Feature Selection Feature Space Image Segmentation Segmentation Result
Unable to display preview. Download preview PDF.
- Haralick, R. M., and Shapiro, L. G. Computer Vision and Robot Vision, vol. I and II, Addison-Wesley, 1992.Google Scholar
- Shahshahani, B. M., and Landgrebe, D. The Effect of Unlabeled Samples in Reducing the Small Sample Size Problem and Mitigating the Hughes Phenomenon. IEEE Transactions on Geoscience and Remote Sensing, 32(5):1087–1095.Google Scholar
- Beardslee, D., and Wertheimer, M., Principles of perceptual organization. Readings in Perception, Van Nostrand, Princeton, NJ, 1958, pp. 115–135.Google Scholar
- Dellepiane, S. The active role of 2D and 3D images: semi-automatic segmentation. Contemporary Perspectives in Three-Dimensional Biomedical Imaging, 1997, ch 7, Roux C. and Coatrieux J.L. Eds., IOS Press.Google Scholar
- Kass, M., Witkin, A., and Terzopoulos, D. Snakes: active contour models. Proc. First Int. Conf. Comp. Vision (1987), pp. 259–268.Google Scholar
- EU Project BREAKIT (INFO 2000) - contact: Giunti Interactive Labs, Genova.Google Scholar