Image Compression Techniques: An Analysis. Evaluation using Higher Order Metrics

      Аннотация: Importance of image compression increases with advancing communication technology. Limited hardware and budget is also important in sending data fast. The amount of data associated with visual information is so large that its storage requires enormous storage capacity and require large capacity and bandwidth, which could be very expensive.Generally images carry three main types of information: redundant, irrelevant, and useful.Redundant information is that which can be reproduced, without loss, from other information contained in the image, for example, low varying background information. Irrelevant information is the part of the information that has enormous details which is beyond the limit of perceptual significance. Useful information is part of the information which is neither redundant nor irrelevant. Decompressed images are usually observed by human beings. Therefore their fidelities are subject to the capabilities and limitations of the human visual system (HVS).Set partitioning in hierarchical trees (SPIHT), is an extension of Shapiro’s embedded zerotrees wavelet. the algorithms allow for progressive transmission require no training,and are of low computational complexity.

Ключевые слова: Image Compression, JPEG, quantization, SPIHT, Higher order metrics