High Efficient Haar Wavelets for Medical Image Compression

In this paper, we proposed an improved high efficient Haar wavelets (HEHW) algorithm to improve the quality of image compression ratio (CR) rate and peak signal to noise ratio for medical imaging. The proposed algorithm starts by partitioning the original image into 2 * 2 submatrices. Then the wavelets transform coefficients obtained by working on the submatrices instead of the rows and on the columns in the original image. We re-compute the resulting coefficients for sub-matrices to obtain the approximation and the sub details of the original image. Then, we calculate statistical thresholds on the details subbands to complete the compression process. The proposed algorithm is applied to five different medical image structures to prove its efficiency using the evaluation factors like CR, peak signal to noise ratio, mean square error and transform time. The comparison between the proposed and the well-known modified Haar method also compared to the results of existing wavelets techniques: like Coiflet, Daubechies, Biorthogonal, Dmeyer, and Symlets to prove the proposed algorithm efficiency.