Data compression techniques

Data compression techniques are developed to increase the efficiency of communication channels and to occupy less space in storage media. This technique helps reducing communication period, enabling more data flow and decreasing the cost of hardware for data storage. Almost in all data transfer and data storage needs, data compression techniques are extensively used. Amongst the many applications, image data compression has great importance due to its high redundant data content. In this study, different image data compression techniques are investigated and compared with each other for their efficiency and effects on image quality. Among the many data compression techniques, the following ones are choosen to be compared with each other: Run-Length Coding, Discrete Fourier Transform, Discrete Cosine Transform, Vector Quantization, Hierarchical Finite-State Vector Quantization and Wavelet Transform techniques. The basic criteria in comparing these techniques are their compression rates versus their signal to noise ratio. This study has revealed that at low compression rates Run Length Coding and in high compression rates Vector Quantization and Wavelet Transform are more efficient than the others. As alternative approaches, techniques which are called hybrid Wavelet Transform with Vector Quantization technique and hybrid Wavelet Transform with Hierarchical Finite-State Vector Quantization are proposed and presented. The results show that these combinations have better performance than the each algorithm individually at high compression ratios.