This PDF file contains the editorial “Guest Editorial: Special Section on Wavelet Applications” for JEI Vol. 7 Issue 04

We present three recent developments in wavelets and subdivision: wavelet-type transforms that map integers to integers, with an application to lossless coding for images; rate-distortion bounds that realize the compression given by nonlinear approximation theorems for a model where wavelet compression outperforms the Karhunen-Loève approach; and smoothness results for irregularly spaced subdivision schemes, related to wavelet compression for irregularly spaced data.

We investigate the use of vector quantizers (VQs) with memory to encode image sequences. A multiscale video coding technique using adaptive finite-state vector quantization (FSVQ) is presented. In this technique, a small codebook (subcodebook) is generated for each input vector from a much larger codebook (supercodebook) by the selection (through a reordering procedure) of a set of appropriate codevectors that is the best representative of the input vector. Therefore, the subcodebook dynamically adapts to the characteristics of the motion-compensated frame difference signal. Several reordering procedures are introduced, and their performance is evaluated. In adaptive FSVQ, two different methods, predefined thresholding and rate-distortion cost optimization, are used to decide between the supercodebook and subcodebook for encoding a given input vector. A cache-based vector quantizer, a form of adaptive FSVQ, is also presented for very-low-bit-rate video coding. An efficient bit-allocation strategy using quadtree decomposition is used with the cache-based VQ to compress the video signal. The proposed video codec outperforms H.263 in terms of the peak signal-to-noise ratio and perceptual quality at very low bit rates, ranging from 5 to 20 kbps. The picture quality of the proposed video codec is a significant improvement over previous codecs, in terms of annoying distortions (blocking artifacts and mosquito noises), and is comparable to that of recently developed wavelet-based video codecs. This similarity in picture quality can be explained by the fact that the proposed video codec uses multiscale segmentation and subsequent variable-rate coding, which are conceptually similar to wavelet-based coding techniques. The simplicity of the encoder and decoder of the proposed codec makes it more suitable than wavelet-based coding for real-time, very-low-bit-rate video applications.

Tactical battlefield surveillance systems will require the transmission of compressed video to utilize the limited communication bandwidth and data capacity of these systems. Any compression techniques used will result in some loss of information. It is important to assess the quality of the output video to determine its performance in aided target recognition applications. The traditional rate of distortion formula is shown by Mallet [S. Mallet, "Understanding wavelet image compression," Proc. SPIE Wavelet Apps. IV 3078, 74–93 (April 1997); "A theory for multiresolution signal decomposition: The wavelet representation," IEEE Trans. Pattern Anal. Mach. Intell. 11, 674–693 (1989)] to be inappropriate for wavelet compression in high compression ratios. The reason is that the histogram changes from all gray scale to a concentration singularity near the origin of very low bit rate such that the discrete approximation of the density function of the histogram is no longer valid. Thus we cannot theoretically predict the distortion due to wavelet compression. Therefore we conduct an empirical investigation to evaluate the spatial and temporal effects of lossy wavelet compression and reconstruction on tactical infrared video. We quantify localized peak signal-to-noise ratio and feature persistence measure measurements and objective assessment techniques developed by the Institute for Telecommunication Sciences, U.S.Department of Commerce, to assess video impairment based on quality measurements. We therefore measure video degradation rather than absolute video quality, which is difficult to quantify.

Accurate object tracking is important in defense applications where an interceptor missile must hone into a target and track it through the pursuit until the strike occurs. The expense associated with an interceptor missile can be reduced through a distributed processing arrangement where the computing platform on which the tracking algorithm is run resides on the ground, and the interceptor need only carry the sensor and communications equipment as part of its electronics complement. In this arrangement, the sensor images are compressed, transmitted to the ground, and decompressed to facilitate real-time downloading of the data over available bandlimited channels. The tracking algorithm is run on a ground-based computer while tracking results are transmitted back to the interceptor as soon as they become available. Compression and transmission in this scenario introduce distortion. If severe, these distortions can lead to erroneous tracking results. As a consequence, tracking algorithms employed for this purpose must be robust to compression distortions. In this paper we introduced a robust object tracking algorithm based on the continuous wavelet transform. The algorithm processes image sequence data on a frame-by-frame basis, implicitly taking advantage of temporal history and spatial frame filtering to reduce the impact of compression artifacts. Test results show that tracking performance can be maintained at low transmission bit rates and can be used reliably in conjunction with many well-known image compression algorithms.

Recent advances in areas of both discrete wavelet transforms and continuous wavelet transforms representing human visual system neural networks have resulted in improved video compression, restoration, and filtering techniques. These software techniques are capable of achieving quality performance in video, and the computational complexity requires a special design hardware called WaveNet to run a real time live video through radio. The brassboard integrated with computers can potentially provide us with many applications, including remote sensors, security systems, and commercial and home video teleconferencing. This paper describes a low cost board to support a video compression, restoration, and filter system in real time processing. The WaveNet board has been optimized for wavelet-based image and video compression and enhancement techniques.

In this paper we introduce wavelet video processing of proximity sensor signals. Proximity sensing is required for a wide range of military and commercial applications, including weapon fuzing, robotics, and automotive collision avoidance. While our proposed method temporarily increases signal dimension, it eventually performs data compression through the extraction of salient signal features. This data compression in turn reduces the necessary complexity of the remaining computational processing. We demonstrate our method of wavelet video processing via the proximity sensing of nearby objects through their Doppler shift. In doing this we perform a continuous wavelet transform on the Doppler signal, after subjecting it to a time-varying window. We then extract signal features from the resulting wavelet video, which we use as input to pattern recognition neural networks. The networks are trained to estimate the time-varying Doppler shift from the extracted features. We test the estimation performance of the networks, using different degrees of nonlinearity in the frequency shift over time and different levels of noise. We give the analytical result that the signal-to-noise enhancement of our proposed method is at least as good as the square root of the number of video frames, although more work is needed to completely quantify this. Real-time wavelet-based video processing and compression technology recently developed under the DOD WAVENET program offers an exciting opportunity to more fully investigate our proposed method.

A wavelet, being a bank of matched filters at a constant fidelity, is appropriate for low bandwidth video communication–compression and is also proposed here to test the video contrast sensitivity surface. This generalization permits us to evaluate the trade-off between the spatial redundancy versus temporal redundancy reductions. This paper addresses the relationship between very low bandwidth video compression and human perception for target discrimination using video.

Video compression strategy consists of a tacit tradeoff of the spatial redundancy, whose spatial singularities are captured efficiently in terms of zero-phase symmetric wavelet transform (WT) subband coefficients, with their temporal redundancy. On the other hand, the redundancy is robust under noisy channel, and furthermore, an over reduction spatially may hurt the temporal correspondence principle yielding a large frame rate reduction. The video error resilient codes are built upon the single frame error resilient codes, which must have known positions partitioned carefully in order to protect against wireless channel errors during transmission of the discrete WT coefficients. These coefficients have been compressed based on the traditional compression scheme using scalar quantization, runlength, and Huffman coding. The discrete WT biorthogonal (7, 9) wavelet is used. The code is illustrated with relatively good performance for (480348038 bits) forward-looking infrared images, compressed at up to 0.09 bits per pixel, and bit error rates at about 1023. Furthermore, in this paper we adopt the single frame error resilient coding and extend to video multiple frames with the help of frame error concealment techniques. An attempt is made to perform decompression error concealment, which is an interpolation from neighborhood in space and time.

With the widespread use of color management systems (CMSs), users are now able to achieve device independent color across different media. However, most of the current CMSs guarantee the same color only if one sees color under a controlled viewing condition. If one sees color under a different viewing condition, the reproduced color does not match the original. The effect of ambient light on the appearance of the color of softcopy images is discussed in this article. In a typical office environment, a computer graphic monitor with a correlated color temperature (CCT) of 9300 K is widely used under an F6 fluorescent light of 4150 K CCT. In such a case, the human visual system is partially adapted to the CRT monitor’s white point and partially to the ambient light. A new adaptation model, S-LMS, is proposed to compensate for the mixed chromatic adaptation. Visual experiments were performed to evaluate the mixed chromatic adaptation. Experimental results indicated that human visual system is 60% adapted to the monitor’s white point and 40% to ambient light when viewing softcopy images.

A general framework is presented for the development of image histogram modification techniques based on models of the human visual system. A new model is then derived from measurements of the nonlinear characteristic of the visual system reported in the psychophysiology literature, leading to the development of a new histogram modification technique. A modified version of a previously proposed technique, Mokrane’s transformation, is introduced and shown to be a special case of this more general technique. Also discussed is another special case, called quadratic hyperbolization. Examples are provided showing that images processed with the new technique may have greater intelligibility than when processed with existing methods of histogram modification.

The desire to print high quality, highlight color documents at production rates has generated the need for a single-pass highlight color laser scanner. We examine a pulsed-imaging, pulse-width-modulated polygon laser scanner that enables multilevel exposure and subsequent development of black and highlight color toner. As currently used in the Xerox 4850 and 4890 digital printers, this laser scanner writes three different exposure levels; zero-level exposure, mid-level exposure and high exposure for black, white, and highlight color, respectively. In this paper we review the functioning of a pulsed-imaging laser scanner and then examine how it is used in a pulse-width-modulated trilevel mode. A first-order linear-systems model is used to examine the specialized video required for achieving uniform line appearance for both black and highlight color. This model further describes how resolution enhancement may be enabled. A complete laser/xerographic model is used to generate simulated output. The electronics that provide the pulse width and position modulated video signals are described, and photomicrographs of sample prints from the Xerox 4850 are shown.

We study and analyze passive imaging arrays and apertures for their performance in the imaging of spatially incoherent source distributions. The concept of difference coarray is used to provide interesting insight into the aperture synthesis process using frequency diversity or wide-band imaging. We also exploit this understanding of wide-band operation to design imaging arrays, and illustrate our results through examples. The proposed methods are also applicable to imaging of coherent distributions of reflectors.