Color Image Compression Using Geometric Wavelets
- Publisher: Journal of Engineering Computers & Applied Sciences
Journal of Engineering Computers & Applied Sciences
(issn: 2319-5606, eissn: 2319-5606)
Applied Sciences; Department of Mathematics | Binary space partition scheme, color image compression, geometric wavelets, piecewise polynomial approximation, sparse geometric representations
arxiv: Computer Science::Computer Vision and Pattern Recognition
acm: ComputingMethodologies_COMPUTERGRAPHICS | Data_CODINGANDINFORMATIONTHEORY | ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION
In this paper, geometric wavelets are considered in the context of color image compression. Geometric wavelet is a recent development in the field of nonlinear piecewise polynomial approximation. The present study extends the geometric wavelet image coding method for monochromatic images to color images. In proposed algorithm, color images are represented by vector valued functions and after that binary space partitioning (BSP) scheme is applied by considering all the three color components simultaneously. The performance of presented color image coder is compared with Naive-EZW (NEZW), Naive-SPIHT (NSPIHT), CZW, SP-CZW and JPEG baseline coder. The first experimentation of vector valued geometric wavelets to color image compression has given promising results.