AbstractsComputer Science

A multiple application card is a smart card that can support newlinedifferent types of applications on the card itself where an individual bearing a newlinesingle card will be able to interact with several service providers The major newlinechallenge faced by the smart card is its limited memory size Compression newlinetechniques decrease image file size by reducing the data needed to represent newlinean image This research focuses on lossy image compression techniques for newlinemulti application smart card It is desirable to provide visually lossless but newlinein fact lossy compression of still images and to reduce the impact of newlinecompression in image retrieval newlineTo achieve the above goals of compression and retrieval newlineinvestigations are carried out using the existing wavelet based image newlinecompression techniques while maintaining a Peak Signal to Noise Ratio newlinePSNR of 40 to maintain visually lossless compression A wavelet filter newlinebank for improving the compression ratio without loss of visual quality and newlinean improved segmentation technique based on active contours using an newlineoptimization technique to improve compression performance are proposed newlinePerformance measure of machine intelligence algorithms and proposed Multi newlineLayer Perceptron Neural Network MLPNN on the compressed images are newlineEvaluated Finally optimization of Levenberg Marquardt L M algorithm newlineusing Artificial Bee Colony ABC to identify system parameters is proposed newline newline%%%appendix p144-151, reference p152-165.