APPLICATIONS OF LMS ALGORITHM FOR IMAGE COMPRESSION AND ENHANCEMENT

Abstract

CHAPTER -- 1 IN I I ODUC I I ON Digital processing of an input sequence to generate an output sequence, in accordance with a prescribed alegarithm, is commonly known as Digital Signal Processing. Depending on the type of the input and the output sequence, the digital signal processing can be classified as one-dimensional or two dimensional signal processing. There has been a major emphasis on the processing of two dimensional data because of its widespread applications in space imagery, Biomedicine, Geophysics and Radio-astronomy. When these two-dimensional signals are images, tha signal processing is termed as Digital Image Processing. The interest in the digital image processing methods stems from the two principal applications areas i.e the improvement of pictorial information for human interpretation and, processing of scene data~for autonomous machine perception [11. One of the first image processing techniques was used for improvement in the digitized newspaper pictures sent by submarine cable in early 1920s. Although the improvements on processing methods fo- transmitted digital pictures continued to be made after the 'first useful application, it took the advent of large scale computers and the space prpgramme, to bring into focus the potential of image processing concepts. The work on using computer techniques for improvement in image obtained from a space probe began at Jet Propulsion Laboratory in 1964, when the pictures of moon transmitted by Ranger 7 were processed by a computer to correct E various types of image distortion inherent in the on-board television camera. The image processing techniques can be classified into four major areas [2] viz. i) Image Digitization ii) Image Enhancement and restoration iii) Image Coding snd compression iv) Image segmentation and representation. Image compression is useful in storage and transmission where the aim is to minimize the storage and bandwidth requirement for transmission. Typically, a compressed image when decoded to reconstruct its original from, is accompanied by distortion. Thus, the efficiency of a compression algorithm is measured by the following parameters. a) Data compression ability b) Distortion c) Implementation complexity with the last factor being the most significant from hardware-implementation point of view. The various techniques for image coding viz, the pulse code Modulation (PCM), Differential Pulse Code (DPCM), transform coding, hybrid coding etc. have been reviewed by Jain [3], Jayant [4] and Netravali et al [5]. Image enhancement attempts to improve the subjective appearance of an image or to give additional information about the object which was not present in the original image. The principal objective of enhancement techniques is to process a given image so that the result is more suitable than the original image for a particular application [6]. Much improvement can be obtained by various filtering techniques, designed to remove the noise added to the original object during the process of construction and digitization or by emphasizing the outline of some salient feature of the object image. Basically, image enhancement is one of optimal filtering with respect to some error measure.