| The geometric moment invariants were applied to the image compression coding and an image compression algorithm based on geometric moment variants was proposed. Compared to the existing image compression algorithms based on discrete cosine transform and wavelet transform, the similarity between adjacent sub-regions rather than correlation among neighboring pixels was used by the proposed algorithmIn order to verify the image compression algorithm based on geometric moment variants is feasible and rational, this issue the similarity between sub-regions was discussed in detail, the implementation steps of the proposed algorithm in application of gray images was listed and the compression results was compared with the compression algorithm based on discrete cosine transform, and the conclusion that in the same compression ratio the proposed compression algorithm had better visual effects was drawn. Finally, the proposed compression algorithm was applied to true color images, the implementation steps were listed in detail and the compression results were shown in this issue.A platform for digital image compression was designed using VC6.0. The platform can compress 8-bit grayscale images and 24-bit true color images by compression based on moment invariants, JPEG compression and fractal image compression, give compression ratio (CR) and peak signal to noise ratio (PSNR) in dialog box, and display the original image and retrieval images in different windows, which is convenient to analysis results and compare compression performance. |
PDF Full Text Request