Modeling and quality assessment of halftoning by error diffusion

Digital halftoning quantizes a graylevel image to one bit per pixel. Halftoning by error diffusion reduces local quantization error by filtering the quantization error in a feedback loop. In this paper, we linearize error diffusion algorithms by modeling the quantizer as a linear gain plus additive noise. We confirm the accuracy of the linear model in three independent ways. Using the linear model, we quantify the two primary effects of error diffusion: edge sharpening and noise shaping. For each effect, we develop an objective measure of its impact on the subjective quality of the halftone. Edge sharpening is proportional to the linear gain, and we give a formula to estimate the gain from a given error filter. In quantifying the noise, we modify the input image to compensate for the sharpening distortion and apply a perceptually weighted signal-to-noise ratio to the residual of the halftone and modified input image. We compute the correlation between the residual and the original image to show when the residual can be considered signal independent. We also compute a tonality measure similar to total harmonic distortion. We use the proposed measures for edge sharpening, noise shaping, and tonality to evaluate the quality of error diffusion algorithms.     

Design of tone-dependent color-error diffusion halftoning systems.

Grayscale error diffusion introduces nonlinear distortion (directional artifacts and false textures), linear distortion (sharpening), and additive noise. Tone-dependent error diffusion (TDED) reduces these artifacts by controlling the diffusion of quantization errors based on the input graylevel. We present an extension of TDED to color. In color-error diffusion, which color to render becomes a major concern in addition to finding optimal dot patterns. We propose a visually meaningful scheme to train input-level (or tone-) dependent color-error filters. Our design approach employs a Neugebauer printer model and a color human visual system model that takes into account spatial considerations in color reproduction. The resulting halftones overcome several traditional error-diffusion artifacts and achieve significantly greater accuracy in color rendition.     

Color error-diffusion halftoning

Grayscale halftoning converts a continuous-tone image (e.g., 8 bits per pixel) to a lower resolution (e.g., 1 bit per pixel) for printing or display. Grayscale halftoning by error diffusion uses feedback to shape the quantization noise into high frequencies where the human visual system (HVS) is least sensitive. In color halftoning, the application of grayscale error-diffusion methods to the individual colorant planes fails to exploit the HVS response to color noise. Ideally the quantization error must be diffused to frequencies and colors, to which the HVS is least sensitive. Further it is desirable for the color quantization to take place in a perceptual space so that the colorant vector selected as the output color is perceptually closest to the color vector being quantized. This article discusses the design principles of color error diffusion that differentiate it from grayscale error diffusion, focusing on color error diffusion halftoning systems using the red, green, and blue (RGB) space for convenience.     

Design and analysis of vector color error diffusion halftoningsystems

Traditional error diffusion halftoning is a high quality method for producing binary images from digital grayscale images. Error diffusion shapes the quantization noise power into the high frequency regions where the human eye is the least sensitive. Error diffusion may be extended to color images by using error filters with matrix-valued coefficients to take into account the correlation among color planes. For vector color error diffusion, we propose three contributions. First, we analyze vector color error diffusion based on a new matrix gain model for the quantizer, which linearizes vector error diffusion. The model predicts the key characteristics of color error diffusion, esp. image sharpening and noise shaping. The proposed model includes linear gain models for the quantizer by Ardalan and Paulos (1987) and by Kite et al. (1997) as special cases. Second, based on our model, we optimize the noise shaping behavior of color error diffusion by designing error filters that are optimum with respect to any given linear spatially-invariant model of the human visual system. Our approach allows the error filter to have matrix-valued coefficients and diffuse quantization error across color channels in an opponent color representation. Thus, the noise is shaped into frequency regions of reduced human color sensitivity. To obtain the optimal filter, we derive a matrix version of the Yule-Walker equations which we solve by using a gradient descent algorithm. Finally, we show that the vector error filter has a parallel implementation as a polyphase filterbank.     

Tone-dependent error diffusion

We present an enhanced error diffusion halftoning algorithm for which the filter weights and the quantizer thresholds vary depending on input pixel value. The weights and thresholds are optimized based on a human visual system model. Based on an analysis of the edge behavior, a tone dependent threshold is designed to reduce edge effects and start-up delay. We also propose an error diffusion system with parallel scan that uses variable weight locations to reduce worms.     

Quantization of accumulated diffused errors in error diffusion.

Due to its high image quality and moderate computational complexity, error diffusion is a popular halftoning algorithm for use with inkjet printers. However, error diffusion is an inherently serial algorithm that requires buffering a full row of accumulated diffused error (ADE) samples. For the best performance when the algorithm is implemented in hardware, the ADE data should be stored on the chip on which the error diffusion algorithm is implemented. However, this may result in an unacceptable hardware cost. In this paper, we examine the use of quantization of the ADE to reduce the amount of data that must be stored. We consider both uniform and nonuniform quantizers. For the nonuniform quantizers, we build on the concept of tone-dependency in error diffusion, by proposing several novel feature-dependent quantizers that yield improved image quality at a given bit rate, compared to memoryless quantizers. The optimal design of these quantizers is coupled with the design of the tone-dependent parameters associated with error diffusion. This is done via a combination of the classical Lloyd-Max algorithm and the training framework for tone-dependent error diffusion. Our results show that 4-bit uniform quantization of the ADE yields the same halftone quality as error diffusion without quantization of the ADE. At rates that vary from 2 to 3 bits per pixel, depending on the selectivity of the feature on which the quantizer depends, the feature-dependent quantizers achieve essentially the same quality as 4-bit uniform quantization.     

On lattice quantization noise

We present several results regarding the properties of a random vector, uniformly distributed over a lattice cell. This random vector is the quantization noise of a lattice quantizer at high resolution, or the noise of a dithered lattice quantizer at all distortion levels. We find that for the optimal lattice quantizers this noise is wide-sense-stationary and white. Any desirable noise spectra may be realized by an appropriate linear transformation (“shaping”) of a lattice quantizer. As the dimension increases, the normalized second moment of the optimal lattice quantizer goes to 1/2πe, and consequently the quantization noise approaches a white Gaussian process in the divergence sense. In entropy-coded dithered quantization, which can be modeled accurately as passing the source through an additive noise channel, this limit behavior implies that for large lattice dimension both the error and the bit rate approach the error and the information rate of an additive white Gaussian noise (AWGN) channel     

A fast, high-quality inverse halftoning algorithm for errordiffused halftones

Halftones and other binary images are difficult to process with causing several degradation. Degradation is greatly reduced if the halftone is inverse halftoned (converted to grayscale) before scaling, sharpening, rotating, or other processing. For error diffused halftones, we present (1) a fast inverse halftoning algorithm and (2) a new multiscale gradient estimator. The inverse halftoning algorithm is based on anisotropic diffusion. It uses the new multiscale gradient estimator to vary the tradeoff between spatial resolution and grayscale resolution at each pixel to obtain a sharp image with a low perceived noise level. Because the algorithm requires fewer than 300 arithmetic operations per pixel and processes 7x7 neighborhoods of halftone pixels, it is well suited for implementation in VLSI and embedded software. We compare the implementation cost, peak signal to noise ratio, and visual quality with other inverse halftoning algorithms.     

JPEG-compliant perceptual coding for a grayscale image printingpipeline

We describe a procedure by which Joint Photographic Experts Group (JPEG) compression may be customized for gray-scale images that are to be compressed before they are scaled, halftoned, and printed. Our technique maintains 100% compatibility with the JPEG standard, and is applicable with all scaling and halftoning methods. The JPEG quantization table is designed using frequency-domain characteristics of the scaling and halftoning operations, as well as the frequency sensitivity of the human visual system. In addition, the Huffman tables are optimized for low-rate coding. Compression artifacts are significantly reduced because they are masked by the halftoning patterns, and pushed into frequency bands where the eye is less sensitive. We describe how the frequency-domain effects of scaling and halftoning may be measured, and how to account for those effects in an iterative design procedure for the JPEG quantization table. We also present experimental results suggesting that the customized JPEG encoder typically maintains "near visually lossless" image quality at rates below 0.5 b/pixel (with reference to the number of pixels in the original image) when it is used with bilinear interpolation and either error diffusion or ordered dithering. Based on these results, we believe that in terms of the achieved bit rate, the performance of our encoder is typically at least 20% better than that of a JPEG encoder using the suggested baseline tables.     

Spurious-Free Dynamic Range of a Uniform Quantizer

Quantization plays an important role in many systems where analog-to-digital conversion and/or digital-to-analog conversion take place. If the quantization error is correlated with the input signal, then the spectrum of the quantization error will contain spurious peaks. Although analytical formulas describing this effect exist, numerical evaluation can take much effort. This brief provides approximations for the spurious-free dynamic range (SFDR) of a uniform quantizer with a single sinusoidal input, with and without additive Gaussian noise. It is shown that the SFDR increases by approximately 8 dB/bit, in case there is no noise. Generalizing this result to multitone inputs results in an additional 2 dB/bit per additional tone. Additive Gaussian noise decorrelates the sinusoid(s) and the quantization error, which results in a dramatic increase in SFDR.     

Predictive Coding of Speech at Low Bit Rates

Predictive coding is a promising approach for speech coding. In this paper, we review the recent work on adaptive predictive coding of speech signals, with particular emphasis on achieving high speech quality at low bit rates (less than 10 kbits/s). Efficient prediction of the redundant structure in speech signals is obviously important for proper functioning of a predictive coder. It is equally important to ensure that the distortion in the coded speech signal be perceptually small. The subjective loudness of quantization noise depends both on the short-time spectrum of the noise and its relation to the short-time spectrum of the Speech signal. The noise in the formant regions is partially masked by the speech signal itself. This masking of quantization noise by speech signal allows one to use low bit rates while maintaining high speech quality. This paper will present generalizations of predictive coding for minimizing subjective distortion in the reconstructed speech signal at the receiver. The quantizer in predictive coders quantizes its input on a sample-by-sample basis. Such sample-by-sample (instantaneous) quantization creates difficulty in realizing an arbitrary noise spectrum, particularly at low bit rates. We will describe a new class of speech coders in this paper which could be considered to be a generalization of the predictive coder. These new coders not only allow one to realize the precise optimum noise spectrum which is crucial to achieving very low bit rates, but also represent the important first step in bridging the gap between waveform coders and vocoders without suffering from their limitations.     

Double-loop sigma-delta modulation with DC input

A discrete-time model having a two-bit (2-b) quantizer is analyzed exactly, and an analytical expression for the quantizer noise sequence is found. Rigorous answers are then provided to two fundamental questions for a double-loop sigma-delta modulation system with DC input: (1) What is the long-term statistical behavior of the internal quantizer noise? (2) How does the asymptotic mean square sigma-delta quantization error vary as a function of the oversampling ratio?     

MPEG－2视频编码的自适应量化器设计

本文在研究ＭＰＥＧ－２ＴＭ５建议的自适应量化策略的基础上，设计了一种新的自适应量化器。以块为基础分析宏块的局部视觉活动特性，并通过综合评价宏块中各块的视觉活动特性来最终决定自适应视觉量化因子。实验结果表明，本文所设计的自适应量化器能均匀分布图像编码主观失真，改善了图像质量，特别是减少了平坦区的块效应，降低了平坦区强边缘的失真。     

Variable bit rate adaptive predictive coder

An adaptive predictive coder providing almost toll quality at 16 kb/s and minimal degradation when the bit rate is lowered to 9.6 kb/s is described. The coder can operate at intermediate bit rates and can also change bit rate on a packet-by-packet basis. Variable bit rate operation is achieved through the use of switched quantization, thus eliminating the need for buffering of the output. A noise shaping filter provides flexible control of the output noise spectrum. The filter, in conjunction with an enhanced way to adapt the quantizer step size, which tries to accommodate the quantization noise feedback, accounts for the toll quality. By quantizing the residue with more than one quantizer, the effective number of bits per sample can be controlled in a deterministic way regardless of the entropy residue. The lower limit of operation is at 9.6 kb/s. Performance of the coder under random bit errors is also presented. It has been found that only at error rates of 10^-2 and higher does the degradation becomes objectionable     

Oversampled Sigma-Delta Modulation

Oversampled sigma-delta modulation has been proposed as a practical implementation for high rate analog-to-digital conversion because of its simplicity and its robustness against circuit imperfections. To date, mathematical developments of the basic properties of such systems have been based either on simplified continuous-time approximate models or on linearized discrete-time models where the quantizer is replaced by an additive white uniform noise source. In this paper, we rigorously derive several basic properties of a simple discrete-time single integrator loop sigma-delta modulator with an accumulate-and-dump demodulator. The derivation does not require any assumptions on the correlation or distribution of the quantizer error, and hence involves no linearization of the nonlinear system, but it does show that when the input is constant, the state sequence of the integrator in the encoder loop can be modeled exactly as a linear system in an appropriate space. Two basic properties are developed: 1) the behavior of the sigma-delta quantizer when driven by a constant input and its relation to uniform quantization, and 2) the rate-distortion tradeoffs between the oversampling ratio and the average mean-squared quantization error.     

Hardcopy image barcodes via block-error diffusion.

Error diffusion halftoning is a popular method of producing frequency modulated (FM) halftones for printing and display. FM halftoning fixes the dot size (e.g., to one pixel in conventional error diffusion) and varies the dot frequency according to the intensity of the original grayscale image. We generalize error diffusion to produce FM halftones with user-controlled dot size and shape by using block quantization and block filtering. As a key application, we show how block-error diffusion may be applied to embed information in hardcopy using dot shape modulation. We enable the encoding and subsequent decoding of information embedded in the hardcopy version of continuous-tone base images. The encoding-decoding process is modeled by robust data transmission through a noisy print-scan channel that is explicitly modeled. We refer to the encoded printed version as an image barcode due to its high information capacity that differentiates it from common hardcopy watermarks. The encoding/halftoning strategy is based on a modified version of block-error diffusion. Encoder stability, image quality versus information capacity tradeoffs, and decoding issues with and without explicit knowledge of the base image are discussed.     

Quantization Noise in ADPCM Systems

The system considered here consists of a differential pulse-code modulator (DPCM) in which the quantizer is replaced by an adaptive quantizer. Adaptation is accomplished by adjusting the stepsize at every sampling instant, depending upon the magnitude of the quantization error. Quantizing noise in adaptive DPCM (ADPCM) systems falls into three categories: granularity, slope overload, and quantizer saturation. Granular noise occurs because only a finite number of levels are available to represent the analog input signal during the encoding process. Slope overload happens when the slope of the input signal increases faster than the adaptive system can follow. Quantizer saturation exists because nonoptimal decisions on step-size adjustments can lead to situations in which quantizer overload occurs without slope overload. The result is an error larger than a purely granular noise analysis would suggest. Equations for the three types of quantizing noise in ADPCM systems are derived, and computer simulations are perforated. For flatand RC-filtered Gaussian input signals, oversampled at various rates, the simulation results agree well with theoretical predictions. Comparisons indicate that the ADPCM can perform better than the best analogous nonadaptive system in terms of signal-to-quantizing-noise ratio. Furthermore, the optimal operating point for the ADPCM is much less sensitive to changes in input signal parameters and system component values than in nonadaptive systems.     

Quantization Based on the Mean-Absolute-Error Criterion

Performance criteria for the design of optimum quantizers are considered. A distance criterion for quantizer input and output probability distribution functions is formulated, and its relationship to the usual distortion criteria is established. The criterion based on absolute error is shown to have unique properties, justifying its use as a performance index for optimum quantization. Numerical results are presented, and the implication for adaptive quantization of the use of this criterion is discussed.     

On optimal quantization of noisy sources

Problems in optimal multidimensional quantization of sources corrupted by noise are addressed. Expressions for the optimum quantizer values and the optimum quantization rule for the weighted squared error distortion measure are found and calculated for the Gaussian signal in additive independent Gaussian noise problem. Some properties of the optimum quantizer, and its relations with the optimal estimator for the general problem, are derived     

Properties of minimum mean squared error block quantizers (Corresp.)

Two results in minimum mean square error quantization theory are presented. The first section gives a simplified derivation of a well-known upper bound to the distortion introduced by ak-dimensional optimum quantizer. It is then shown that an optimum multidimensional quantizer preserves the mean vector of the input and that the mean square quantization error is given by the sum of the component variances of the input minus the sum of the variances of the output.