Some remarks on very high-rate image compression and its impact on security image data subjective evaluation

The paper deals with selected problems of image transmission through extremely narrow communication channels using very high compression rates. Such conditions are very common in the security technology field. The same problem relates to image retrieval systems when limited storage capacity is available. As an example of a typical security scenario, a seven-segment display was chosen and compressed at various rates. The subjective identification threshold was consequently evaluated. Four recent compression techniques were involved: JPEG (DCT), LuraWave (WT), JPEG2000 (WT) and fractals. Relevant compression rates were evaluated for the identification threshold and the standard test image LENA was processed under identical conditions. The results are discussed.