Multispectral band image compression using adaptive wavelet transform - Tetrolet transform

The proposed work develops an efficient multi spectral band imagery compression technique using adaptive haar wavelet transform, called tetrolet transform. Geometrical features are most important prominent factor in multispectral image processing. But existing compression algorithms are fail to preserve geometrical features at high compression ratio, which implicates the visual cognitive effects around features of reconstructed image. In tetrolet transform, the determined orthonormal basis functions are adapted to geometrical features of the image. By using the filter bank algorithm, the tetrolet transform coefficients of input image are obtained. Thresholding and encoding steps are used to achieve compression of an image. The reconstruction is done by using decoding and inverse tetrolet transform. This proposed method produces improved performance than existing extended shearlet based compression technique. The simulation results show that the proposed method produces better performance than extended shearlet based compression technique by the factors of peak signal to noise ratio (PSNR), compression ratio(CR) and bits per pixel (bpp).