Natural Image Correction by Iterative Projections to Eigenspace Constructed in Normalized Image Space

Image correction is discussed for realizing both effective object recognition and realistiic image-based rendering. Three image normalizations are compared in relation with the linear subspaces and eigenspaces, and we conclude that the normalization by LI-norm, which normalizes the total sum of intensities, is the best for our pwyoses. Based on noise analysis in the normalized image space(NIS), an image correction algorithm is constructed, which is accomplished by iterative projections along with corrections of an image to an eigenspace in NH. Experimental results show that the proposed method works well for natural images which include various kinds of noise shadows, refiections and occlusions. The proposed method provides a feasible solution to the object recognition based on the illumination cone /Z?]. The technique can also be eztended to face detection of unknown person and registration/recognition using eigenfaces.