Image Compression - Bezier and Splines in Image Processing and Machine Vision - page 102

Table 4.5. Evaluation of reconstructed Lena images.

Components of F v Lena image

For input For output JPEG result 1 JPEG result 2

MSE 0 96.953 129.638 142.574

PSNR ∞ 28.265 27.003 26.590

ρ xy 1.0 0.977 0.992 0.989

H (1)

2.529 1.909 2.639 2.528

H (2)

0.0021 0.0009 0.00007 0.00046

IQI 27.616 35.700 30.197 30.372

FD 2.619 2.620 2.624 2.593

Table 4.6. Evaluation of reconstructed Girl images.
Components of F v

Girl image
For input For output JPEG result 1 JPEG result 2

MSE 0 52.848 37.236 38.497

PSNR ∞ 31.457 32.421 32.27

ρ xy 1.0 0.987 0.991 0.991

H (1) 1.956 1.591 2.268 2.264

H (2)

0.043 0.018 0.144 0.197

IQI 85.952 115.263 90.619 89.553

FD 2.607 2.577 2.531 2.529

Table 4.7. Some results on Lena image due to increase in spatial resolution.
Image size Article Principle of Compression PSNR in

coding ratio db

512 × 512 [139] vector quantization 12.30 29.95

256 × 256 [33] sketch based 5.30 N.M

256 × 256 [61] fractal 11.85 30.58

512 × 512 [61] fractal 36.78 30.71

512 × 512 [107] fractal 40.00 30.20

512 × 512 [145] segmentation using 31.00 N.M

morphology 20.00 N.M

512 × 512 [141] block prediction 30.76 32.78

512 × 512 [160] region based 41.00 26.56

fractal

512 × 512 [108] fractal 44.00 29.10

512 × 512 [159] fractal 44.44 29.10

256 × 256 [75] fractal 10.60 30.72

512 × 512 [52] wavelet based 65.60 29.90

fractal