Abstract:
Images play a cardinal role in innumerable fields of life. Image compression is nothing but reduction of an image size in bytes by compromising on the quality of original image to a tolerable level. In present era, image compression is preferred by wavelet transform, which is the most in demand “time-frequency” transformation. Wavelet transforms are used due to their intrinsic property that they are redundant and shift invariant. The wavelet transform surpassed the discrete cosine transform (DCT) and its ancestors due to its unique quality that defined the image both in frequency and time domains, while DCT defined it with sine and cosine waves. In low bit rates, artifacts are caused in DCT because it tries to distribute numerous bits to approximations furthermore, a couple of bits are assigned to fine details, whereas DWT has no artifacts. Embedded encoding has witnessed a remarkable progress during the last two decades. With discrete wavelet transform as their basis and concatenation with entropy encodings, have paved the way towards the compression optimization. This paper is an endeavor to present a survey on the above mentioned encoding techniques by underlining their facets and drawbacks in detail. It has been discussed that as to how SPIHT has outperformed EZW not only by addressing few of its inherent shortcomings but also exhibited improvements in the basic parameters of PSNR and Compression Ratios. On the other hand EBCOT has brought improvements by preserving the edges lost by SPIHT.
