Image Coding Based On Optical Wavelet Transform

Optical wavelet transform is the combination of optical information processing and wavelet theory. It integrates many advantages of optical science and wavelet transform while providing advantageous basis for high-speed information processing. Optical wavelet transform for image data compression is one of the frontiers. Realizing image data compression on optical system will have important theoretic meaning and practical value for images'storage and transmission. The fundamental theory of wavelet was described and the effect on entropy for several factors in wavelet transform was studied in this dissertation. Then the methods of image compression after wavelet transform and the optical system's characteristics were analyzed in detail. And nearly lossless compression and lossy compression of optical wavelet coefficients were proposed.The variation of image entropy before and after wavelet decomposition, the optimal number of wavelet decomposition layer, the effect on the entropy by wavelet basis and image's frequency components etc. were studied by a large number of simulation experiments. It is verified that to get the minimal entropy, generally a three-layer decomposition should be adapted rather than higher orders. The result by biorthogonal wavelet decomposition is better than by orthogonal one. And it's true that the higher of the vanishing moment, the better the result.The requirement for wavelet bases in optical wavelet transform is introduced and the methods of image compression with wavelet transform were analyzed in detail. Considering the above experimental results and the requirements of wavelet bases in image compression, the choice of appropriate wavelet bases for the optical system was described. Based on traditional static image compression algorithms, the characteristics of wavelet coefficients and the development of coding methods were introduced. Especially the commonly used embedded coding algorithms were analyzed in detail. And the advantages and disadvantages of these methods were compared.Image data compression based on optical wavelet transform was achieved here. First, characteristics of optical system and optical wavelet transform coefficients were analyzed. According to these characteristics, simulation was did on several quantified and coding methods. Based on the results, the dead zone uniform quantization and EBCOT were chose. And the corresponding experiments confirmed the validity of these methods. In addition, from analyzing the endurance to quantity noise of optical system, an important conclusion was achieved which sparked the ideas of nearly lossless compression. It is described as: so long as the noise caused by quantizing and coding is less than the noise produced by the system itself, the quality of the decoding image would not be affected greatly. Experimental results showed that, compared with the image reconstructed directly, the nearly lossless reconstructed image's PSNR was only 0.1dB lower, and there was no significant difference visually.