In this paper, a new encoding approach is proposed to control the JPEG2000 encoding in order to reach a desired perceptual quality. The new method is based on a vision model that incorporates various masking effects of human visual perception and a perceptual distortion metric that takes spatial and spectral summation of individual quantization errors into account. Compared with the conventional rate-based distortion minimization JPEG2000 encoding, the new method provides a way to generate consistent quality images at a lower bit rate.

The spatial standard observer (SSO) was developed in response to a need for a simple, practical tool for measurement of visibility and discriminability of spatial patterns. The SSO is a highly simplified model of human spatial vision, based on data collected in a large cooperative multi-lab project known as ModelFest. It incorporates only a few essential components, such as a local contrast transformation, contrast sensitivity junction, local masking, and local pooling. The SSO may be useful in a wide variety of applications, such as evaluating vision from unmanned aerial vehicles, measuring visibility of damage to aircraft and to the shuttle arbiter, predicting outcomes of corrective laser eye surgery, inspection of displays during the manufacturing process, estimation of the quality of compressed digital video, evaluation of legibility of text, and predicting discriminability of icons or symbols in a graphical user interface.

In this paper, a new JPEG2000-compliant encoding approach is proposed to control the JPEG2000 encoding in order to achieve a desired perceptual quality. Our method is based on a vision model that incorporates various masking effects of the human visual perception and on a perceptual distortion metric that takes spatial and spectral summation of individual quantization errors into account. Compared with the conventional JPEG2000 rate-based distortion minimization encoding, our method provides a way to generate consistent quality images with lower bit rate.

Video quality metrics are intended to replace human evaluation with evaluation by machine. To accurately simulate human judgement, they must include some aspects of the human visual system. In this paper we present a class of low-complexity video quality metrics based on the standard spatial observer (SSO). In these metrics, the basic SSO model is improved with several additional features from the current human vision models. To evaluate the metrics, we make use of the data set recently produced by the Video Quality Experts Group (VQEG), which consists of subjective ratings of 160 samples of digital video covering a wide range of quality. For each metric we examine the correlation between its predictions and the subjective ratings. The results show that SSO-based models with local masking obtain the same degree of accuracy as the best metric considered by VQEG (P5), and significantly better correlations than the other VQEG models. The results suggest that local masking is a key feature to improve the correlation of the basic SSO model.

The discrete wavelet transform (DWT) decomposes an image into bands that vary in spatial frequency and orientation. It is widely used for image compression. Measures of the visibility of DWT quantization errors are required to achieve optimal compression. Uniform quantization of a single band of coefficients results in an artifact that we call DWT uniform quantization noise; it is the sum of a lattice of random amplitude basis functions of the corresponding DWT synthesis filter. We measured visual detection thresholds for samples of DWT uniform quantization noise in Y, Cb, and Cr color channels. The spatial frequency of a wavelet is r 2-lambda, where r is display visual resolution in pixels/degree, and lambda is the wavelet level. Thresholds increase rapidly with wavelet spatial frequency. Thresholds also increase from Y to Cr to Cb, and with orientation from lowpass to horizontal/vertical to diagonal. We construct a mathematical model for DWT noise detection thresholds that is a function of level, orientation, and display visual resolution. This allows calculation of a "perceptually lossless" quantization matrix for which all errors are in theory below the visual threshold. The model may also be used as the basis for adaptive quantization schemes.

An extension to the JPEG standard (ISO/IEC DIS 10918-3) allows spatial adaptive coding of still images. As with baseline JPEG coding, one quantization matrix applies to an entire image channel, but in addition the user may specify a multiplier for each 8/spl times/8 block, which multiplies the quantization matrix, yielding the new matrix for that block. MPEG 1 and 2 use much the same scheme, except there the multiplier changes only on macroblock boundaries. We propose a method for perceptual optimization of the set of multipliers. We compute the perceptual error for each block based upon the DCT quantization error adjusted according to the contrast sensitivity, light adaptation, and contrast masking, and pick the set of multipliers which yield maximally flat perceptual error over the blocks of the image. We investigate the bit rate savings due to this adaptive coding scheme and the relative importance of the different sorts of masking on adaptive coding.

The discrete cosine transform (DCT) is a standard method by which images may be compressed. Initially, the image is divided into 8/spl times/8-pixel blocks, each of which is then transformed into its DCT, which is an 8/spl times/8 matrix of coefficients. One image-dependent factor influencing the detectability of DCT basis functions is contrast masking. The authors present data which indicate that sensitivity to a particular coefficient's quantization error is affected by the magnitudes of other coefficients. They propose a revision of Model O to account for between-coefficient contrast masking.< >

The authors have examined the variation in visibility of single DCT basis functions as a function of display visual resolution. They have shown that the existing model (Ahumada and Peterson, 1992; and Peterson et al., 1993) accommodates resolutions of 16, 32, and 64 pixels/degree, provided that one parameter, the peak sensitivity s0, is allowed to vary. Variations in this parameter are to some extent consistent with spatial summation, although sensitivity is lower at the lowest resolution than summation would predict. Practical DCT quantization matrices must take into account both the visibility of single basis functions, and the spatial pooling of artifacts from block to block. Peterson et al. (1993) showed that to a first approximation this pooling is consistent with probability summation. If one considers two images of equivalent size in degrees, but visual resolutions differing by a factor of two, then the sensitivity to individual artifacts would be lower by 4/sup 1/4/ in the higher resolution image due to the smaller block size in degrees, but higher by 4/sup 1/4/ in the same image due to the greater number of blocks. Thus the same matrix should be used with both. The point of the illustration is that the overall gain of the best quantization matrix must take into account both display resolution and image size.< >

The JPEG baseline standard for image compression employs a block discrete cosine transform (DCT) and uniform quantization. For a monochrome image, a single quantization matrix is allowed, while for a color image, distinct matrices are allowed for each color channel. We describe a method, called DCTune, for the design of color quantization matrices that is based on a model of the visibility of quantization artifacts. The model describes the artifact visibility as a function of DCT frequency, color channel, and display resolution and brightness. The model also describes the summation of artifacts over space and frequency, and masking of artifacts by the image itself. The DCTune matrices are different from the de facto JPEG matrices, and appear to provide superior visual quality at equal bit-rates.>

A custom quantization matrix tailored to a particular image is designed by an image-dependent perceptual method incorporating solutions to the problems of luminance and contrast masking, error pooling and quality selectability.>

The authors calculate the spatiotemporal power spectrum of 14 image sequences in order to determine the degree to which the spectra are separable in space and time and to assess the validity of the commonly used exponential correlation model. They expand the spectrum by a singular value decomposition into a sum of separable terms and define an index of spatiotemporal separability. as the fraction of the signal energy that can be represented by the first (largest) separable term. All spectra were found to be highly separable with an index of separability above 0.98. The power spectra of the sequences were well fit by a separable model, which corresponds to a product of exponential autocorrelation functions separable in space and time.>