Application of the source reduction method with multiwavelets to image separation
Akira Morimoto, Ryuichi Ashino, Takeshi Mandai · 2012
The blind image source separation problem is to estimate the number of the unknown source images and to separate the original images from superpositions of source images. The coefficients of the superposition are also unknown. Our proposed source reduction method has a potential to overcome the difficulty of previously proposed methods. In this paper, we give an algorithm of the method with continuous multiwavelet transform for the image separation. Various numerical experiments using the proposed algorithm ensure that our proposed source reduction method performs well for image separation. An example of numerical experiments is presented.