Cognitive Sub-Band Replacement DWT based image compression scheme

High quality color images transmitted wirelessly should exploit minimum resources. Hence designing algorithms that are adaptive in nature has become imperative, in order to achieve the growing demands of multi-media communication. This paper implements a unique image compression technique known as `Cognitive Sub-Band Replacement Discrete Wavelet Transform (CSR-DWT)´. The designed algorithm performs on the fundamental principle of cognitive radio. By using energy thresholding technique on DWT sub-bands, the cognitive band is determined and then replaced by other information thereby creating a unique textured image. This allows the algorithm to achieve a desirable compression ratio up-to 45-60 %. Furthermore this technique possesses the reversible property, that faithfully retrieves original color image from the textured image. It also proves its efficiency without hampering the quality of the image and exhibiting Peak Signal to Noise Ratio (PSNR) of 29-32 dB with correspondingly very low Mean Square Error (MSE) value of 17-19 dB.