International Journal of Electrical and Electronics Research ISSN 2348-6988 (online) Vol. 8, Issue 3, pp: (1-9), Month: July - September 2020, Available at: www.researchpublish.com
DEVELOPMENT OF FRACTAL IMAGE CODING TECHNIQUE FOR MEDICAL IMAGES D. V. Shobhana Priscilla* * Sr. Assistant Professor, Dept of EIE, VNR Vignana Jyothi Institute of Engineering and Technology, Hyderabad. shobana_dv@vnrvjiet.in
Abstract: The domain pool design is one of the dominant issues, which affects the coding performance of the fractal compression. In this paper a new method called Block Averaging Method is used to design an effective domain pool. The key feature in this method is that the redundancy between the adjacent domain blocks is reduced. The generated domain pools are much more efficient than those generated by the conventional fractal coding schemes and thus providing improved coding performance. The proposed iteration-free fractal codec method provides high decoding speed and excellent image quality apart from improving the compression ratio and reducing the bit rate. In addition a Hybrid Model is also proposed which will still more reduce the computational time during the coding process. Keywords: Block Average, Domain Pool, Fractal Image Compression, Iteration-Free. 1.
INTRODUCTION
The storage and transmission of medical images has made compression a necessity. In order to transmit a large number of medical images with less memory usage, the use of compression techniques have gained very much importance in our day-to-day life. The fractal image coding is different in which the image is made up of copies of properly transformed parts(domain pool) of itself. The reduction in the redundancy between the domain blocks in the domain pool is very important in fractal coding as the domain pool is the storehouse for the fractal codes by which the image can be reconstructed without any error failing which the reconstructed image will be different from the original image. Arnaud E. Jacquin [1] proposed an independent and novel approach to image coding based on fractal theory of iterated transformation. The main characteristics of this approach are that i) it relies on the assumption that the image redundancy can be efficiently exploited through self-transformability on a block wise basis and ii) It approximates an original Image by a fractal Image. The feasibility of the design of an expert algorithm for the determination of optimal fractal block coding system parameters which would enable the coder to meet specific image fidelity and bit rate requirements for given type of images to encode is to be investigated. More over the coding and decoding phase of this method is still iterative and hence it involves high computational complexity which affects the memory requirement and also increases the computation time significantly. Hsuan T. Chang et al [6], proposed the Linde, Buzo, and Gray(LBG) method of designing the domain pool in the conventional fractal-coding scheme. Moreover the training process requires iteration to obtain the minimum quantization error. The larger the codebook size, the bigger the iteration number. However it considers the encoding process only and it is not possible to obtain the identical codebook on the decoding phase unless an offline transmission is made. The other option is to apply Linde, Buzo, and Gray (LBG) to generate the domain pool in the decoding phase as well which will further prolong the decoding process.
Page | 1 Research Publish Journals