John O. Limb

Measuring the Speed of Moving Objects from Television Signals

Very simple techniques for estimating the speed of a moving object from a television signal are described. They assume that frame storage is available. One technique requires little more than two threshold circuits and two counters. The accuracy of one technique is better than ± 10 percent, quite sufficient to enable the measure to be used as a control signal for the efficient coding of television type signals.

On the Design of Quantizers for DPCM Coders: A Functional Relationship Between Visibility, Probability and Masking

Visibility functions measure the relative visibility of noise added to a picture at those points where some measure of local activity exceeds a given threshold. The functions are obtained from a series of subjective experiments and vary with the content of the picture. Visibility functions have been used to design quantizing characteristics for DPCM coding of monochrome and color signals and for three-dimensional transform coding. We consider an alternative approach to determining the visibility function that obviates the need for repeated picture-dependent subjective tests. The visibility function is assumed to consist of two parts, a picture-dependent component and viewer-dependent component (referred to as the masking function). The visibility function may be approximated by the quotient of a probability density function raised to a power and the masking function. The role of probability is found to be weaker where the viewer has more opportunity to scrutinize the picture.

Plateau Coding of the Chrominance Component of Color Picture Signals

Plateau coding is a method for efficiently coding the chromatic component of color television signals. The signal is divided into regions of approximately constant chromaticity which are transmitted by sending one set of chrominance values for the whole region. Since most changes in chrominance are accompanied by luminance changes, the luminance signal (which must be transmitted quite accurately) is used at the transmitter and receiver to indicate changes in the two chrominance signals; thus no addresses need be sent to define the boundaries of the region at the receiver. Simulations have shown that plateau coding can give pictures of high quality for chrominance bit-rates in the range 0.25-0.5 bits per luminance sample, assuming straightforward pulse-code modulation (PCM) coding of the chrominance amplitudes.

A Picture-Coding Algorithm for the Merli Scan

A method is proposed and evaluated for coding a notched scan that jumps back and forth between adjacent scan lines of a normal raster. Alternate points in the line are coded using the normal differential quantizer algorithm while the remaining points use a nonlinear predictor proposed by Graham. The advantage of the algorithm over normal differential quantization is that the prediction strategy is improved while the advantage over other two-dimensional strategies is that the effect of transmission errors is confined to the line pair being processed and does not spread vertically. For coarse coding there is a significant improvement in picture quality when compared with normal differential quantization while for high-quality pictures there is a small reduction in entropy. The effect of transmission errors is investigated and shown to be similar to the effect with normal differential quantization.

Digital Coding of Color Picturephone Signals by Element-Differential Quantization

An element differential quantizer has been used to encode three baseband components of a color Picturephone® type signal in real time. The first part of this paper is concerned with an investigation of the extra bits required to transmit chrominance information over that required for a high-quality luminance signal (coded to 4 bits per picture element). We have found that a total allocation of 1 bit per picture element to the chrominance signals leads to a high-quality color display. The second part of the paper concerns an investigation of a more efficient coding format for color sigaals. In this regard, we have determined that the color signal can be packaged into a 6.3-Mbit/s rate by allocating 12 levels for the luminance component, 6 levels for one chrominance component, and 4 levels for the other chrominance component. Only one chrominance component is transmitted each line and the missing component is obtained by line averaging. The best results were obtained by coding chrominance signals that were matrixed such that their color axes lay between the I and Q and color difference axes. A scheme is suggested for simply combining this type of coder with an analog color signal that has the chrominance information compressed into the blanking portion of the signal.

Spatiotemporal characteristics of thresholds adjacent to a luminance edge.

Spatiotemporal incremental and decremental thresholds were measured for a thin vertical line (target) positioned adjacent to a briefly presented vertical edge. A significant difference between the stimuli used here and those used in previous studies is the background level against which the edge was presented. Here the edge was formed by briefly decreasing the luminance of the left side of a light background. This novel condition was compared with the more usual condition in which the edge is formed by briefly increasing the luminance of the right half of a dark background. In a further test the buildup of threshold after an edge was switched on was also measured. When the target was presented on the side of the edge that remained fixed in luminance, a small but reliable threshold change adjacent to the edge was measured. The effect was much larger for measurements made on the side of the edge that changed in luminance; however, the spread was comparable for the two conditions. For the target presented on the light side of the edge, decremental thresholds were much larger than incremental thresholds. This is attributed, at least in part, to the different types of tasks required of the observer. Maximum temporal threshold elevation occurs at or just before (i.e., less than 16 msec) 0 asynchrony between edge and target. The results are interpreted at a qualitative level as supporting a receptive-field type of model in which the edge, at various spatiotemporal locations relative to the target, inhibits or excites activity at the receptive field centered on the target.

Visual Perception Applied To The Encoding Of Pictures

This overview begins by emphasizing the importance of the human observer to the process of efficiently encoding picture material. By way of example, two aspects of threshold vision that bear importantly on the coding problem are discussed: the threshold of a perturbation presented against a plain background and threshold adjacent to a luminance step. Methods for incorporating threshold information into the coding operation are briefly described.

Communications in the Automated Office - Guest Editor's Prologue

The scope of this Special Issue is communication in its most general sense, encompassing communication from person to person, between people and machines, and from machine to machine.

Real-time extraction of bubble chamber tracks using a single vidicon

Abstract Bubble Chamber pictures show many undesired tracks and background in addition to the tracks of the desired significant event. Settles et al. have described a technique for optical tagging of an event by adding a darkfield photograph taken before significant bubble growth to a later brightfield photograph. We describe a system to cancel out all picture detail except for the wanted tracks by using a single vidicon tube as the storage device. In the first exposure, polarized light is imaged on the vidicon after passing through a Ronchi grating placed at a focal plane. Thus half of the target is exposed in a series of vertical stripes. The second exposure uses light polarized orthogonally to the first exposure and is deflected after passing through the Ronchi grating so as to expose the previously occluded stripes on the target. The target is then scanned orthogonally to the stripes; by subtracting the picture contained in one set of stripes from the contained in the other set, only the differences between the two images remains. A simulation was conducted using continuously presented background of one polarization and background plus tracks of the other polarization. The test showed that the added tracks were easily resolved, even though they were not readily discernible by visual inspection prior to subtraction.