248 
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
Figure 6. Decompressed images with bit rates at 0.5 and 0.2. 
(a)is the original image, (b) is the local-enlarged image of (a), 
(c) is the wavelet deconstructed image at the bit rate 0.5, (d) is 
local-enlarged image of (c), (e)is the WBCT deconstructed 
image at the bit rate 0.5, (f) is the local-enlarged image of (e), 
(g) is the wavelet deconstructed image at the bit rate 0.2, (h) is 
the local-enlarged image of (g), (i) is the WBCT deconstructed 
image at the bit rate 0.2, (j) is the local-enlarged image of (i). 
From the images above, we can get to the conclusion that the 
image compressed by WBCT preserves more details than the 
wavelet method. 
3.2 The image quality assessment by SSIM 
After image decompression, we use the SSIM method to 
evaluate the quality of the reconstructed images. And the result 
comparing is shown in Table. 1. 
Compression Rate 
SSIM 
wavelet 
WBCT 
16 
0.8445 
0.8757 
40 
0.7405 
0.7552 
Tabled The SSIM value of the reconstructed images (the 
original image without compression has an SSIM 1). 4 
4. CONCLUSIONS AND FURTHER WORKS 
In this paper we show the design and the implementation of a 
novel directional detail preserving image coder. The stages of 
the coder are: discrete wavelet transformation, directional filter 
deconstruction, bitplane coding, image compression and 
decompression. The possibility of coding images with 
directional detail preserving offers a wide range of applications 
in several industries in which this process is imperative such as 
medicine, mobile devices and face recognition. The classical 
objective measures such as PSNR are not appropriated to 
measure the goodness of this coder, this is due to the goal of the 
coder is not to improve the measures but preserve directional 
features. 
There’s still other work needs to be done. For example, the 
choosing of the wavelet base and the directional filter that can 
best dig out the advantage of WBCT (KE Li and HUANG 
Lian-qing, 2005). And the ImageZip2.0 compression system 
uses integer wavelet coding method which is disadvantage to 
our method because the WBCT coefficients are double float 
type. The loss of data precision makes the PSNR lower than the 
wavelet transformation. So it is necessary to find a suitable float 
encode method to improve our WBCT compression system. 
REFERENCES 
D. Schilling and P. C. March 2001, Cosman, "Feature- 
Preserving Image Coding for Very Low Bit Rates", 
Proceedings of the IEEE Data Compression Conference (DCC), 
Snowbird, Utah, U. S. A., pp. 103 - 112. 
D. Schilling and P. C. December 2003, Cosman, "Preserving 
Step Edges in Low Bit Rate Progressive Image Compression", 
IEEE Transactions on Image Processing, vol. 12, no. 12, pp. 
1473 - 1484. 
FeiPeng Li, 2003, “Remote Sensing Satellite Image 
Compression”, Doctoral Dissertation, TN911, 10486, Wuhan 
University, China. 
K. R. Namuduri and V. N. November 2003, Ramaswamy, 
"Feature Preserving Image Compression", Pattern Recognition 
Letters, vol. 24, no. 15, pp. 2161 - 2776. 
KE Li, HUANG Lian-qing, 2005, Choice of waveletbase in 
real-time compression for remote sensing image. Journal, 
OPTICAL TECHNIQUE, 36(1), 77-83 
LiYu Lin, 2007, “Research on Application of Contourlet 
Transform in Remote Sense Image Processing”, Doctoral 
Dissertation, TN911, 10486, Wuhan University, China 
M. N. Do and M. December 2005 Vetterli, “The Contourlet 
Transform: An Efficient Directional Multiresolution Image 
Representation”, IEEE Transactions on Image Processing, vol. 
14, no. 12,, pp. 2091 -2106. 
O. O. Vergara Villegas, R. Pinto Elias and V. G September 
2006,. Cruz Sánchez, “Feature Preserving Image Compression: 
A Survey”, Proceedings of the Electronics, Robotics and 
Automotive Mechanics Conference (CERMA), vol. 2, 
Cuernavaca, Morelos México, pp. 35 - 40. 
Vivien Chappelier, Christine de Beaulieu, “IMAGE CODING 
WITH ITERATED CONTOURLET AND WAVELET 
TRANSFORMS”, 2004 International Conference on Image 
Processing (ICIP). 
Zhou Wang, Alan Conrad Bovik, Hamid Rahim Sheikh and 
Eero P.Simoncelli, “Image Quality Assessment: From Error 
Visibility to Structural Similarity”, IEEE Transactions on 
Image Processing, vol. 13, no. 4, April. 2005.