Paper Title
Based On Quaternion Matrix Concept Vector Sparse Demonstration Of Color Image
Abstract
In old sparse image model considering color image as a scalar, it derives each color channels independently.
Using quaternion matrix concept in our project we developed a vector sparse demonstration for color image. Some of the
image processing tasks are denoising, inpainting and super-resolution. In our project we considering color image as a
quaternion matrix. In our project we developed K- QSVD algorithm. K-QSVD stands for quaternion singular value
decomposition and K means group of similar for QSVD. In quaternion matrix color image channels are changed to
orthogonal color space. The developed sparse model is more efficient than the present sparse model. The developed sparse
model successfully overcome the any distortion in the image. and it is a powerful tool in the color image processing.
Efficient model of the image representation is sparse demonstration.
Keywords— Vector sparse representation, quaternion matrix analysis, color image, dictionary learning, K-QSVD.