K. Yun, S. Cho, C. Ahn, and S. Lee (Korea)
Image coding, region classification, wavelet transform
We propose a novel image coding method, which efficiently uses the relationship between the properties of a spatial image and its wavelet transform. Firstly, an original image is decomposed into several layers by wavelet transform, and simultaneously decomposed into 2n2n blocks. Each image is classified into three image types according to the standard deviation of its blocks. And then each block is categorized into two regions by the threshold that is different according to the image types, i.e., Significant Activity Region (SAR) and Insignificant Activity Region (IAR). The IAR in spatial domain does not only appear as zero regions in wavelet domain like LH, HL and HH but also gives little influence to the quality of the reconstructed image. On the other hand, the SAR is related to the significant coefficients that affect to image reconstruction. Simulation results show that the proposed coding method has better performance than the EZW and SPIHT in terms of image quality and transmitted bit-rate. Also, it can be applied to the application areas that are necessary of the progressive transmission of the image data.
Important Links:
Go Back
