Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

The reconstructed images of 512 × 512 Lena with range size 8 × 8 for BFIC scheme and BSFIC scheme, respectively.

More »

Fig 1 Expand

Table 1.

PSNR/dB of the reconstructed images in BFIC scheme and BSFIC scheme.

More »

Table 1 Expand

Fig 2.

The generating process of the fast sparse fractal image compression (FSFIC) scheme.

More »

Fig 2 Expand

Fig 3.

The relations of different values of k1 with the reconstructed image quality, the compression ratio, and the encoding time for 512 × 512 images Lena, baboon, and pepper, respectively.

The range size is 8 × 8, k2 is the number of all the domain blocks in the domain pool, and the average pixel-based square error e = 25.

More »

Fig 3 Expand

Fig 4.

The relations of different values of k2 with the reconstructed image quality, the compression ratio, and the encoding time for 512 × 512 images Lena, baboon, and pepper, respectively.

The range size is 8 × 8, k1 = 4, and the average pixel-based square error e = 25.

More »

Fig 4 Expand

Fig 5.

The relations of different values of the average pixel-based square error e with the reconstructed image quality, the compression ratio, and the encoding time for 512 × 512 images Lena, baboon, and pepper, respectively.

The range size is 8 × 8, k1 = 4, and k2 is the number of all the domain blocks in the domain pool.

More »

Fig 5 Expand

Fig 6.

The optimum cases of the reconstructed image quality for different compression ratios with fixing k2 as the number of all the domain blocks in the domain pool.

More »

Fig 6 Expand

Fig 7.

The comparisons among these SFIC schemes with different accelerating technologies on 512 × 512 image Lena with sparse number 4.

These accelerating technologies include the APCC-based classification and sorting method (the proposed FSFIC scheme), the APCC-based only sorting method (only-sorting), the polar angle and NRMS-based method (Kovacs), and the hog feature-based method. Because the encoding time of the HOG-based method is always more than 1.5s in our experiment, it doesn’t appear in this figure.

More »

Fig 7 Expand

Fig 8.

Three examples to show the advantage of convergence of the proposed FSFIC scheme.

The images in the first line are the reconstructed images in the quadtree FIC scheme, and the images in the second line are the reconstructed images in the proposed FSFIC scheme.

More »

Fig 8 Expand