Aiping Yang

All phase biorthogonal transform and its application in JPEG-like image compression

This paper proposes new concepts of the all phase biorthogonal transform (APBT) and the dual biorthogonal basis vectors. In the light of all phase digital filtering theory, three kinds of all phase biorthogonal transforms based on the Walsh transform (WT), the discrete cosine transform (DCT) and the inverse discrete cosine transform (IDCT) are proposed. The matrices of APBT based on WT, DCT and IDCT are deduced, which can be used in image compression instead of the conventional DCT. Compared with DCT-based JPEG (DCT-JPEG) image compression algorithm at the same bit rates, the PSNR and visual quality of the reconstructed images using these transforms are approximate to DCT, outgoing DCT-JPEG at low bit rates especially. But the advantage is that the quantization table is simplified and the transform coefficients can be quantized uniformly. Therefore, the computing time becomes shorter and the hardware implementation easier.

Shape Adaptive All Phase Biorthogonal Transform and Its Application in Image Coding

Abstract —In this paper shape adaptive all phase biorthogonal transform (SA-APBT) for coding arbitrarily shaped image segments is proposed. The proposed transform can be used in region-based image compression instead of the shape adaptive discrete cosine transform (SA-DCT). Region-based image coding method is made up of three procedures: image segmentation, contour coding and texture coding. An image is divided into region-of-interest (ROI) and background area by image segmentation. Both areas are encoded and decoded independently. Experimental results on test images show that compared with SA-DCT at the same bit rates, the PSNRs and visual effects of reconstructed images based on SA-APBT are comparable to the one based on SA-DCT especially at low bit rates. And the SA-APBT transform coefficients are quantized uniformly, while the SA-DCT transform coefficients are quantized according to different quantization steps in the quantization table. So the quantization based on SA-APBT is simpler and computational complexity is reduced.

Image Coding Algorithm Based on All Phase Walsh Biorthogonal Transform

This paper proposes new concepts of the all phase Walsh biorthogonal transform (APWBT) and the dual biorthogonal basis vectors based on the Walsh-Hadamard transform. The matrices of APWBT are deduced. They can be used in image compression instead of the discrete cosine transform (DCT). Compared with DCT in JPEG image compression algorithm, the PSNRs of the reconstructed images using the transform are the same as DCTs at the same bit rates, but the advantage is that the quantization is very simple. The transform coefficients can be quantized uniformly. Therefore the computational complexity is reduced and it is much easier to realize with the hardware.

BINARY TREE IMAGE CODING ALGORITHM BASED ON NON-SEPARABLE WAVELET TRANSFORM VIA LIFTING SCHEME

In recent years, image coding based on wavelet transform has made rapid progress. In this paper, quincunx lifting scheme in wavelet transform is introduced and all phase interpolation filter banks which can be used in the lifting scheme for prediction and update are designed. Based on the basic idea of set partitioning in hierarchical trees (SPIHT) algorithm, the binary tree image coding algorithm is proposed. Just like SPIHT, the encoding algorithms can be stopped at any compressed file size or let run until the compressed file is a representation of a nearly lossless image. The experimental results on test images show that compared with SPIHT algorithm, the PSNRs of the proposed algorithm are superior by about 0.5 dB at the same bit rates and the subjective quality of reconstructed images is also better.

An Improved JPEG Compression Algorithm Based on Sloped-Facet Model of Image Segmentation

This paper proposed an improved JPEG compression algorithm based on Haralick sloped-facet model. Different areas after segmentation were taken into account in this algorithm and different areas are compressed in different proportion. The experiment results show that the quality of reconstructed images is better than the traditional JPEG compression algorithm at the same bit rate about 0.3 dB.

Image deblurring based on wavelet and neural network

A new method of image deblurring based on wavelet decomposition and RBF (Radial Basis Functions) neural network is proposed. Traditional algorithms of image deburring have some limits, such as assumption of generalized stationary process and calculating complexity etc in their application. Based on the data of degraded image and original image as trained samples of the RBF network, the similarly degraded image may be restored through recording the reverse process. Combining wavelet transform with neural network, through wavelet decomposition of degraded image and original image, images of different low frequency resolution may be obtained; thus, multi-resolution image pyramids of the two sample images can be constructed. The RBF network structure may be optimized by levels through the sample images trained from the top to the bottom of the pyramid, and the weights obtained from the upper level were used as the original weights of the next level. The image restoration thus can be conducted by the eventually obtained optimized network. Simulation results show that the proposed method is easily achieved, the convergence is accelerated and the ideal deblurred result is realized.

Video compression based on wavelet transform and DBMA with motion compensation

In video compression, it can reduce the redundant information efficiently by motion estimation (ME) and motion compensation (MC), and less code is used to encode as much information as possible. In this paper, I frame encoding adopts wavelet transform and set partitioning in hierarchical trees (SPIHT) algorithm; for P frames, each frame sets the reconstructed frame of its previous frame as a reference frame, and then P frames proceed to code with ME and MC. In the step of ME, trading off between accuracy and computational complexity, adaptive fast search (AFS) algorithm combining with nodal search-based deformable block matching algorithm (NS-DBMA) is adopted to search for the matched block; and then in the MC process, wavelet transform combining with zerotree entropy (ZTE) algorithm is adopted according to the characteristics of residual image data. Meanwhile rate control is carried out in ZTE algorithm. Experimental results show that the proposed algorithm performs well for video sequences with complex motion and rich details, and reduces blocking artifacts obviously. Index Terms—video compression, motion estimation (ME), motion compensation (MC), wavelet transform, nodal search- based deformable block matching algorithm (NS-DBMA), adaptive fast search (AFS)

Construction of Wavelet Transform Matrices with Symmetric Boundary-Extension

In order to realize perfect reconstruction, finite-length signals must be extended in some way before they can be transformed. This paper first proposes a method of constructing wavelets decomposition and reconstruction matrices of wavelet coefficient without iterative calculation with symmetric boundary-extension of finite length signal, and correspondingly gives the base vectors and base graphs of 9/7 wavelet as an example. While the edge effect can be evidently reduced compared with the periodic boundary-extension method, the matrices constructed with symmetric boundary-extension of signal in this paper have better performance and can be widely used for image-partition processing based on wavelet method

Image coding based on image segmentation and shape adaptive all phase biorthogonal transform

Based on the shape adaptive discrete cosine transform (SA-DCT) and its application in image coding, this paper proposes a new image coding algorithm based on image segmentation and shape adaptive all phase biorthogonal transform (SA-APBT). In this paper, an image is divided into two parts, the region-of-interest (ROI) and the background area, which can be encoded separately. Intra macroblocks (all pixels are located in ROI or background region) are processed with APBT; while marginal macroblocks (part of pixels are located in ROI) are processed with SA-APBT. Experimental results are obtained with the test images. It can be concluded that the coding performance of the proposed algorithm is better than that of conventional algorithms. Both the objective quality and subjective effect are improved.

Binary tree image compression algorithm based on wavelet transform via lifting scheme

In this paper, all phase interpolation filter banks were designed and binary tree wavelet coding algorithm was proposed based on the basic idea of lifting scheme and SPIHT algorithm. The encoding algorithms can be stopped at any compressed file size or let run until the compressed file is a representation of a nearly lossless image. The experiment results on test images show that compared with SPIHT algorithm, the PSNR of the proposed algorithm is superior at the same bit rates about 0.5 dB and the subjective quality of reconstructed images is better too.