DD296591A5 - ARRANGEMENT FOR PROCESSING A WIDE VIDEO SIGNAL - Google Patents

Abstract

The invention relates to an arrangement for processing a widescreen video signal from which a widescreen with a Bildseitenverhaeltnis, which is greater than a standard Seitenseitenverhaeltnis arises. The wide-screen video signal has a central field signal component having a given aspect ratio and a side field signal component having low-frequency and high-frequency components coded by processes different from each other. The arrangement can be applied to a wide-screen television system which is compatible with conventional television receivers whose aspect ratio is smaller than the aspect ratio of wide-screen images. The arrangement comprises, according to the invention, a device for decoding the low-frequency side field signal component, a device for decoding the high-frequency side field signal component and a device for modifying the aspect ratio of the central field signal component. There is also provided means for combining decoded low and high frequency side field signal components with the central field signal component having the modified aspect ratio. Fig. 13 {arrangement; Processing; Widescreen video signal; Bildseitenverhaeltnis; Central field signal component; Side panel signal component; decoding; Modifier; Combiner; low frequency component; high-frequency component}

Description

16 pages drawings

Field of application of the invention

The invention relates to an arrangement for processing a widescreen video signal, from which a widescreen with an aspect ratio greater than a standard aspect ratio, arises, wherein the widescreen video signal having a central field signal component with a given aspect ratio and a side field signal component with low-frequency and high-frequency components, with Help were coded by mutually different processes. The arrangement according to the invention can be applied to a wide-screen television system which is compatible with conventional television receivers whose aspect ratio is smaller than the aspect ratio of wide-screen images.

Characteristic of the known state of the art

A conventional television receiver such. For example, a receiver tuned to the NTSC broadcast standard used in the US and elsewhere has an aspect ratio (ratio of width to height of the reproduced image) of 4: 3.

Recently, however, there is interest in higher aspect ratios for television receivers, e.g. As the ratios 2: 1, 16: 9 or 5: 3. Such higher aspect ratios correspond to the visual field of the human eye rather than the 4: 3 aspect ratio of a conventional television receiver.

Special attention is paid to video signals for a picture aspect ratio of 5: 3, since this value corresponds to the aspect ratio of cinema films. Such video signals can then be transmitted and received without clipping the image information. However, widescreen television systems (also called "wide-screen" systems), which simply transmit higher aspect ratio signals than conventional systems, are incompatible with conventional aspect ratio receivers, making widespread adoption of widescreen systems difficult.

Thus, there is a need for a widescreen system that is compatible with conventional television receivers. ^

Object of the invention

The aim of the invention is to increase the picture quality of reproduced television pictures over the entire display area and to allow a higher aspect ratio than in conventional television receivers, which corresponds better to the visual field of the human eye.

-2- 296 591 Presentation of the Essence of the Invention

The invention has for its object to provide an arrangement for processing a widescreen video signal, from which a widescreen image aspect ratio, which is greater than a standard aspect ratio, arises. The wide-screen video signal should have a central field signal component having a given aspect ratio and a side field signal component having low-frequency and high-frequency components encoded by mutually different processes.

According to the invention the object is achieved in that the arrangements comprises four devices, wherein the first means for decoding the low-frequency side field signal component, the second means for decoding the high-frequency side field signal component and the third means for modifying the aspect ratio of the central field signal component is used. There is further provided means for combining decoded low and high frequency side field signal components with the central field signal component having the modified aspect ratio. Here, the given aspect ratio of the central field signal component is a standard aspect ratio of approximately 4: 3.

According to the invention, the arrangement is characterized in that the modifying means temporally compresses the aspect ratio of the central field component and that the low-frequency component is coded by the temporal compression of the low-frequency component into the horizontal Bildabtabtastgebiet. According to a further feature of the invention, the high frequency component of the sideband signal component is encoded by modulating an alternating subcarrier with the high frequency side field component, the alternating subcarrier having a frequency different from the chrominance subcarrier frequency.

embodiments

FIG. 1 is a block diagram of an encoder for a compatible widescreen television system in FIG

Accordance with the principles of the present invention;

Figs. 2 to 5 show signal waveforms useful in understanding the operation of the system of Fig. 1;

Fig.6bis10,

12u.12abis12d: show aspects of the system of FIG. 1 in more detail; Figs. 10a to 10c: illustrate aspects of the filter arrangement associated with a filter network shown in Fig. 10; 11 illustrates an amplitude-frequency response curve along with a feature of the system of FIG

FIG. 1; Fig. 13: shows a block diagram of a part of a wide-screen television receiver including a

inventive decoding apparatus; and Fig. 14: shows a portion of the decoder of Fig. 14 in greater detail.

A brief overview of the system to be described will be useful. A system capable of transmitting television pictures having a large aspect ratio of e.g. B. 5: 3 via a standard channel such. B. an NTSC channel is, should achieve a high-quality image reproduction by means of a widescreen receiver, while observable quality deterioration in a standard reproduction with a screen aspect ratio of 4: 3 are largely reduced or eliminated. The application of signal compression techniques to the side panels of an image uses the horizontal overscan area when playing back with a standard NTSC television receiver, but may result in image resolution losses in the side panel areas of a redesigned widescreen. A temporal compression leads to the extension of the frequency range. Therefore, only low frequency components survived processing in a standard television channel having a narrower bandwidth compared to the bandwidth required for a wide-screen signal. Thus, if the compressed side panels of a compatible widescreen signal are expanded in a widescreen receiver, there will be a significant difference between the resolution or high frequency content of the central portion of a rendered widescreen and the side panels if no steps are taken to avoid this effect.

The figures and the related description illustrate a system for developing a widescreen signal that can be processed over a standard NTSC channel. The system also allows a widescreen receiver to provide a widescreen with good picture quality over the entire display area.

As can be seen by the encoder of Figure 1, the application of spatial compression allows the compression of the low frequency side panel information into the horizontal overscan area of a standard NTSC signal. The high frequency side panel information is spectrally divided over the image transmission channel with the standard NTSC signal, in a manner that is transparent to a standard receiver through the use of an alternate subcarrier modulation scheme.

Before discussing the compatible widescreen encoding system of FIG. 1, reference is made to the signal waveforms A and B in FIG. Signal A is a 5: 3 widescreen signal that has been converted to a standard 4: 3 aspect ratio NTSC compatible signal, as indicated by signal B. The wide picture signal A includes left and right side panel sections, each including TS intervals, which are typically of equal duration, and a central panel section, which includes a TC interval.

The wide picture signal A was converted into an NTSC signal B by fully compressing certain side panel information into the horizontal oversampling areas, including the time intervals TO. The NTSC signal comprises an active line interval TA (with a duration of approximately 52.5 microseconds) comprising the oversampling intervals TO, a playback time interval TD containing the image information to be displayed, and a total horizontal line interval TH having a duration of approximately 63.556 microseconds. The intervals TA and TH are the same for the wide-screen signal and the NTSC signal.

It has been found that nearly all home connected television receivers have an oversampling interval that is at least 4% of the total active line time TA, ie 2% oversampling on the left and right sides. At a sampling frequency of 4 χ f _x (where f _{sc is} the frequency of the color subcarrier), each horizontal line interval contains 910 pixels (pixels), of which 754 represent the active horizontal line image information to be displayed.

A widescreen camera 10 according to Figure 1 provides a widescreen color signal with R, G and B components and an aspect ratio of 5: 3 in this example. A widescreen camera is substantially identical to the standard NTSC camera except that a widescreen camera has a larger aspect ratio and a larger image bandwidth. The image width of a widescreen camera is, among other factors, proportional to the product of its aspect ratio and the total number of lines per field (image). When a constant scanning speed of the wide-screen camera is assumed, enlargement of the aspect ratio causes a corresponding increase in the image bandwidth and horizontal compression of the image information when the color image signal is reproduced by a standard television receiver having a 4: 3 aspect ratio. For these reasons, it is necessary to modify the widescreen signal for full NTSC compatibility.

The color image signal processed by the coding system according to FIG. 1 contains both luminance and chrominance signal components. The luminance and chrominance signals contain both low and high frequency information, which will be referred to as "lows" in the following discussion. The broad bandwidth wide color image signals coming from camera 10 are matrixed in unit 12 to derive the luminance component Y and the color difference signal components I and Q from the R, G and B color signals. The widescreen Y, -I and -Q signals are sampled at a fourfold chrominance subcarrier frequency (f χ f _x ) and individually by separate analog-to-digital converters (ADC) in an ADC unit 14 from analog to digital (binary) form implemented. These signals are then individually filtered by separate horizontal low pass filters (LPF) in a filter unit 16 to produce filtered YF, IF and QF signals. These signals each have the waveform A shown in Fig. 2. The luminance signal YF is limited in bandwidth by filters 16 to CEF x 4.2 MHz or approximately 5 MHz, CEF being the expansion factor of the central field. This is necessary so that, following the subsequent time expansion, the bandwidth of the center field signal, as will be discussed, is reduced to 4.2MHz, the bandwidth of an NTSC image signal. For a similar reason, the signals IF and QF are limited by filter 16 to a bandwidth of CEF χ 50OkHz or approximately 60OkHz. The filter unit 16 therefore has a luminance cutoff frequency (cutoff frequency) of approximately 5.0 MHz, and an I and Q cutoff frequency of approximately 60 kHz.

The bandwidths of the Y, I and Q filters of the unit 16 are related to the central field expansion factor, which in turn is a function of the difference between the width of an image reproduced by a wide-screen receiver and the width of an image reproduced by a standard receiver. The image width of a widescreen with an aspect ratio of 5: 3 is 1.25 times larger than the image width of a standard screen with a 4: 3 aspect ratio. This factor of 1.25 is a preliminary centroid expansion factor that must be regulated to account for the oversampling area of a standard receiver as well as the intended low overlap of the central field-to-side-field boundaries, as will be discussed. These considerations dictate a central field expansion factor of 1.19.

The broadband signals coming from the filter unit 16 are processed by a side field central field signal separator and a processor 18 to produce three sets of output signals: YE, IE, and QE; YO, IO and QO; and LH, RH; IH and QH. The first two signal groups (YE, IE, QE and YO, IO, QO) are processed in a first channel which develops a signal containing a full bandwidth central field component as well as low frequency side field luminance information forced into horizontal overscan regions. The third set of signals (LH, RH, IH and QH) is processed in a second channel which develops a signal containing low frequency side panel information. When the output signals of the two channels are combined, an NTSC-compatible wide-screen signal having an aspect ratio of the reproduced picture of 4: 3 is produced. Details of the circuits containing the unit 18 are shown and discussed in connection with FIGS. 6, 7 and 8.

The signals YE, IE and QE contain complete central field information and have the same format, as indicated in Figure 3 by the signal YE. The following is a brief description of how a signal YE is derived from a YF signal. The wide-screen broadband signal YF coming from the unit 16 contains the pixels 1 to 754 which appear during the active line interval of the board image signal containing page and central field information. The wideband central field information (pixels 75 to 680) is extracted by time demultiplexing as the central field luminance signal YC.

Signal YC is time stretched by the center field expansion factor 1.19 (i.e., 5.0MHz-4.2 MHz) to produce an NTSC-compatible central field signal YE. On Grurid's time expansion factor of 1.19, the YE signal has an NTSC compatible bandwidth (0 to 4.2MHz). The YE-SignaJ occupies the image reproduction interval TD (Figure 2) between the oversampling areas TO (pixels 1 to 14 and 741 to 754). The signals IE and QE are developed from the signals IF and QF and are processed similarly to the YE signal.

The signals Y0, IO and Q0 provide the low frequency side field information ("lows") which are inserted into the left and right horizontal oversampling areas The signals Y0, IO and Q0 have the same format as indicated by the Y0 signal in FIG 3, the YO signal is derived from the YF signal as follows: The wide-field signal YF includes left side panel information including pixels 1 through 84 and right side panel information including pixels 671 through 754. As will be discussed, the YF signal for the generation of a low-frequency luminance signal having a bandwidth of 0 to 70OkHz, a signal from which a left and a right low-frequency side-field signal are extracted by a time demultiplexing method (signal YL 'in Fig. 3), low-pass filtered.

The low-frequency luminance signal YL 'is compressed with compressed low-frequency information to generate a low side field signal YO in the oversampling areas including pixels 1 to 14 and 741 to 754. The compressed low frequency side panel information has an increased bandwidth that is proportional to the amount of temporal compression. The signals IO and QO are developed from the signals IF and QF and are processed in a similar way as signal YO.

The signals YE, IE, QE and YO, IO and QO are output from a side field central field signal combiner 28, e.g. B. by a time multiplexer for producing the signals YN, IN and QN combined with an NTSC-compatible bandwidth and an aspect ratio of 4: 3. These signals are of the form shown by signal YN in FIG. The combining means 28 also enables suitable signal delays for the compensation of the transit times of the signals to be combined. Such compensation signal delays are also provided elsewhere in the system where it is necessary to compensate for signal propagation delays.

The chrominance signals IN and QN are quadrature modulated by a modulator 30 at the NTSC chrominance subcarrier frequency of 3.58 MHz on an intermediate carrier SC. As will be described in connection with Figure 9, the modulated signal, before being applied to the chrominance signal input of an NTSC encoder 36, is converted into vertical (V) and temporal by means of a 2-D filter (a two-dimensional filter) 32 (T) Dimension lowpass filtered. The luminance signal YN, before being applied to the luminance input of the encoder 35, is filtered by a 3-D filter 34 (a three-dimensional filter) in the horizontal (H), vertical (V) and temporal (T) dimensions with a band rejection filter filtered. The filtering of the luminance signal QN and the chrominance color difference signals IN and QN serve to ensure that the luminance-chrominance crosstalk is significantly reduced following the subsequent NTSC encoding. The luminance filter 34 filters the luminance signal in the spectral range in which the high-frequency luminance side-field information is modulated, such as a band rejection filter, as will be discussed.

The multi-dimensional space-time filters, such as the 3-D filter (HVT filter) 34 and the 2-D filters 32 and 46 (VT filters) have the structure as illustrated in FIG. Fig. 10 specifically illustrates a vertical time filter {VT filter) which can operate by setting the weighting coefficients a 1 to a9 such as a VT bandpass filter, a VT bandstop filter or a VT low-pass filter. The table of Figure 10a illustrates the weighting coefficients for the VT bandpass filter and VT bandstop filter embodiments used in the disclosed system.

An HVT notch filter, such as filter 34 of Figure 1, includes the combination of a horizontal lowpass filter 1020 and a VT bandstop filter 1021, as shown in Figure 10b. The HVT bandpass filters as included in the decoding system of Figure 13 comprise the combination of a horizontal bandpass filter 1030 and a VT bandpass filter 1031, as shown in Figure 10c.

In the HVT notch filter of Figure 10b, the horizontal low pass filter 1020 has a given cutoff frequency and provides a filtered low frequency signal component. This signal is subtractively combined in a combiner 1023 to generate a high frequency signal component having a delayed type of input signal coming from the delay unit 1022. The low-frequency component is subjected to a one-frame delay by means of a network 1024 prior to application to an additional combiner 1025. The high frequency component is filtered by the VT bandstop filter 1021 before being applied to the adder 1025 for providing an output signal filtered by an HVT bandstop filter. The VT filter 1021 has the VT notch filter coefficients indicated in FIG. 10 a.

An HVT bandpass filter as included in the decoder of FIG. 13 is shown in FIG. 10c with a horizontal bandpass filter 1030 having a given cutoff frequency and a VT bandpass filter 1031 having the VT bandpass filter coefficient as shown in FIG Table according to FIG 10a indicated, is connected in cascade. The filter of FIG. 10 includes a plurality of cascaded storage units (M) 1010a to 1010h for providing successive signal delays at the respective taps t2 to t9, as well as for providing a total filter delay. The signals transmitted by the taps are applied to one input of the multipliers 1012 a to 1012 i, respectively. Another input of each multiplier receives a prescribed weighting coefficient a 1 to a9 depending on the type of filtering process to be performed. The nature of the filtering process also dictates the delays brought about by the memory units 1010a-1010h. The horizontal filters have pixel storage elements so that the total filter delay is less than the time interval of a horizontal image line (1H).

In the case of the vertical filters, only line storage elements are used and in the case of the time filters only image storage elements are used. Thus, an HVT filter (3-D filter) comprises a combination of pixel memory element (<1 H), line memory element (1 H) and image memory element (> 1 H), whereas a VT filter only the last one includes two types of memory elements. The evaluated tapped (mutually delayed) signals of the elements 1012 a to 1012 i are combined to produce a filtered output in an adder 1015.

Such filters are non-recursive finite impulse response (FIR) filters. The type of delay produced by the memory elements depends on the type of filtered signal and the amount of crossover, which in this example can be tolerated between luminance, chrominance and high frequency side panel information. The sharpness of the filter-lock characteristic is increased by increasing the number of memory elements connected in cascade. The HVT bandstop filter 34 according to FIG. 1 has the structure according to FIG. 10 b and removes the upwardly moving diagonal frequency components from the luminance signal YN. These frequency components are similar in appearance to the chrominance subcarrier components and are removed to create a hole in the frequency spectrum / into which high frequency chrominance field information and high frequency luminance side panel information are inserted. The removal of the upwardly moving diagonal frequency components from the luminance signal YN does not visibly degrade a reproduced image because it has been found that the human eye is substantially insensitive to these frequency components. Filter 34 has a cut-off frequency of approximately 1.5 MHz, so that the individual luminance vertical information is not deteriorated.

The VT bandpass filter 32 reduces the chrominance bandwidth so that the modulated chrominance side panel information can be inserted into the hole created by the filter 34 in the luminance spectrum. The filter 32 reduces the vertical and temporal resolution of the chrominance information so that the static and moving edges are slightly blurred, but this effect has no or little consequence due to the insensitivity of the human eye to such an effect. An output signal C / SL from the encoder 36 contains NTSC-compatible information to be reproduced as derived from the central field of the wide-screen signal. This output signal also includes compressed low frequency side panel information (both luminance and chrominance information) derived from the side panels of the widescreen signal and located in the left and right horizontal oversampling areas, which are not visible to the viewer of an NTSC receiver image. The crowded low frequency side panel information in the oversampling area forms part of the side panel information for widescreen display. The other component, the high frequency side panel information, is generated as follows:

The processor 18 forms the signals LH (high frequency left side field luminance information), RH (high frequency right side field luminance information), TH (high frequency! -Information) and QH (high frequency Q information) in the signal processing channel for the high side field information. These signals are illustrated in FIGS. 4 and 5. Figures 6, 7 and 8 illustrate the apparatus for developing these signals. In Fig. 4, a signal YH 'derived from the wide-field signal YF includes high-frequency left side panel information including the left side panel pixels 1 to 84 and high-frequency right side panel information including the right side panel pixels 671 to 754. The high frequency information in this example includes a bandwidth of 700 kHz to 5.0 MHz. For each horizontal line, the high frequency left sidebase information component between pixels 1 and 84 of signal YH 'is time stretched (which reduces its bandwidth) by a page expansion factor and mapped in the central field occupied by pixels 85 through 670, such that one component LH (Figure 4) is generated by side panel information.

At the same time, for each horizontal line, the high frequency right side panel information component between pixel 671 and pixel 754 of signal YH 'is also time expanded and in the central field occupied by pixels 85-670 to generate another simultaneous component RH (Figure 4) of the side panel information added. The concurrent signals RH and LH have a reduced bandwidth, which is the ratio between the stretched sidebase width and the original sidebar width, due to the sidewall expansion factor (6,96). The signals LH and RH are temporally multiplexed with the signals IH and QH by means of a luminance chrominance multiplexer 42 to simultaneously generate the high frequency side field signal components X and Z as shown in FIG. The signal component X is generated by inserting the high frequency left luminance component LH (pixels 85 to 670) between the high frequency left and right side field information of the color difference signal IH. Likewise, at the same time, the signal component Z is generated by inserting the high-frequency right luminance component RH (pixels 85 to 760) between the high-frequency left and right side-field information of the color difference signal QH. The signals X and Z, which contain high-frequency side field information, each have a bandwidth of 0 to 70OkHz and are quadrature modulated by means of a quadrature modulator 43 to a horizontally synchronized alternating subcarrier signal ASC. The frequency of the alternate subcarrier signal ASC is selected to ensure adequate separation (e.g., 20 to 30dB) of the page and central field information, and to insignificantly interfere with the image reproduced by a standard NTSC receiver. In this embodiment, the signal ASC has a frequency of 2.368 MHz.

The frequency of 2.368 MHz selected for the alternate subcarrier signal ASC is an interline frequency, an odd multiple of one-half of the horizontal line frequency, ie 301 χ f _H / 2. This alternating subcarrier frequency produces a fine virtually imperceptible interference fringe pattern which does not affect the quality of the reproduced image as compared to the grave "moving stripes" of the interface pattern which would be caused by the non-interlaced intercarrier frequency The intercarrier frequency of 2.368 MHz is advantageously substantially As shown in Fig. 11, as a result, the modulated high frequency side panel information between the vertical detail band and the chrominance frequency band has a bandwidth of ± 700 kHz, to simplify this illustration Figure 11 does not illustrate the actual total luminance frequency spectrum, which extends to 4.2 MHz and, as is known, is interleaved with the chrominance frequency spectrum.

The square modulation advantageously allows the simultaneous transmission of two narrowband signals. The stretching of the high frequency side field signals results in their bandwidth being reduced in accordance with the narrowband requirements of square modulation. The more the bandwidth is reduced, the less likely it is that interference will occur between the carrier signal and the modulated signal. It should also be noted that the described multiple time utilization method of the high-frequency luminance and chrominance side-field information for the generation of X and Z signals prior to quadrature modulation advantageously requires only one subcarrier instead of two. Furthermore, since the DC component of the side panel information is crowded into the oversampling area, the energy of the modulated signal and, therefore, the potential interference of the modulated signal are greatly reduced.

To reduce the likelihood of the disturbance produced by the quadrature modulated signal, the signal from the modulator 43 is attenuated by the attenuator 44, which causes a signal gain of 0.25, before being transmitted in the vertical and temporal planes (VT plane). is filtered by the bandpass filter 46 along the diagonal axes. The effect of the attenuator 44 has been found to be to reduce the visibility of certain types of disturbances caused by uncorrected modulated high frequency side panel information when viewing the image on a standard NTSC receiver.

The attenuation achieved by the network 44 can also be brought about by adjusting the weighting factors of the filter 46. A filtered quadrature modulated output signal SH from the filter 46 containing the high frequency side panel information is combined with the C / SL signal in combiner 40 to produce an NTSC compatible widescreen NTSC signal. The NTSC signal is converted into an analog form by a digital-to-analog converter (DAC) 54 prior to application to the RF modulator and transmission network 55 for transmission via an antenna 56. The encoded NTSC compatible widescreen signal broadcast by the antenna 56 is intended to be received by both NTSC receivers and widescreen receivers as illustrated in FIG. However, prior to the discussion of Figure 13, reference is made to Figures 6 through 9 and 12, which illustrate in greater detail certain portions of the coding system of Figure 1.

FIG. 6 illustrates an apparatus associated with the processor 18 of FIG. 1 and designed to develop signals YE, YO, LH and RH from the wideband wide-field signal YF. The YF signal is horizontally filtered by a low pass filter 610 having a cut-off frequency of 700 kHz to produce a low-frequency luminance signal YL which is applied to an input of a subtractive combiner 612. The YF signal is applied to another input of combiner 612 and a time demultiplexer 616 after being delayed by a real time 614 to compensate for the signal processing delay of filter 610. The combination of the delayed signal YF and the filtered signal YL produces a high-frequency luminance signal! YH at the output of the combiner 612..

The delayed signal YF and the signals YH and YL are applied to separate inputs of the demultiplexer 616, which includes the demultiplexing units (DEMUX units) 618, 620 and 621 for processing the signals YF, YH and YL, respectively. The details of the demultiplexer 616 will be discussed in connection with FIG. As shown in FIGS. 3 and 4, the demultiplexing units 618, 620 and 621 respectively derive the full bandwidth central field signal YC, the high side field signal YH 'and the low side field signal YL'.

The YC signal is time stretched by the time expander 622 to produce the YE signal, whereas the time stretchers 624 and 626 stretch the signal YH 'to produce the signals LH and RH, respectively. Signal YC is time stretched with a central expansion factor sufficient to leave space for the left and right horizontal oversampling areas. The central expansion factor (1.19) is the ratio of the intended width of the YE signal (pixels 15 to 740) to the width of the YC signal (pixels 75 to 680), as shown in FIG. The signal YH 'is stretched with a page expansion factor to produce the signal LH. The page expansion factor (6.97) is the ratio of the intended width of the signal LH (pixels 85 to 670) to the width of the left side panel component of signal YH '(pixels 1 to 84) as shown in FIG. The signal RH is generated by a similar method.

The signal YL 'is compressed by a page compression factor from the time compressor 628 to generate the YO signal. The side compression factor (0.166) is the ratio of the intended width of signal YO (eg, left pixels 1 to 14) to the width of the corresponding portion of signal YL '(eg, left pixels 1 to 84) as shown in FIG 3 is shown. The time stretchers 622, 624, and 626 and the time compressors 622, 624, and 626 and the time compressor 628 are of the type described and illustrated with reference to FIG.

The signals IE, IH and IO or OE, OH and QO are formed from the signals IF and QF in a manner similar to that by which the signals YE, YH 'and YO are formed by the apparatus of FIG. In this connection, reference is made to FIG. 7, which illustrates the apparatus for the formation of the signals IE, IH and IO from the signal IF. The signals QE, QH and QO are formed from the signal QF in a similar manner.

After the wideband wideband IF signal is delayed by the unit 714, it is applied to the demultiplexer 716 and also subtractively combined to produce the high frequency signal IH 'with the low frequency signal IL coming from a low pass filter 710 in a subtractive combiner 712 , The delayed signal IF and the signals IH 'and IL are demultiplexed by the demultiplexing units 718, 720 and 721, respectively, to which the demultiplexing apparatus 716 belongs to generate the signals IC, IH and IL'. To generate the signal IE, the signal IC is time-stretched by a time expander 722, and to generate the signal IO, the signal IL 'is time-compressed by a compressor 728. Signal IC is stretched with a central expansion factor similar to that used for signal YC as discussed, and signal IL 'is compressed with a side compression factor similar to that used for signal YL' as also discussed.

In connection with the arrangements according to Figures 6 and 7 it must be mentioned that z. For example, in FIG. 6, the filtering of the input signal instead of after-the-demultiplexer 616 advantageously avoids the unwanted transient signal edge phenomena in the output signals LH, RH, and YO. Specifically, the demultiplexer 616 produces outputs having sharp, well-defined output transitions that are distorted by filtering the output signals from the demultiplexer 616, e.g. B. (smeared) would.

In Figure 8, a demultiplexer 816, as can be used for the apparatus 616 of Figure 6 and apparatus 716 of Figure 7, is illustrated. The apparatus according to FIG. 8 is illustrated in connection with the demultiplexer 616 in FIG. The input signal YF contains 754 pixels which map the image information. The pixels 1 to 84 map the left field, the pixels 671 to 754 represent the right field, and the pixels 75 to 680 map the center field slightly overlapping the left and right side panels. The signals IF and QF have a similar overlap. As will be discussed, it has been found that such field overlap facilitates the combination (splicing) of the center panel with the side panels at the receiver and thus substantially eliminates boundary line formation. The demultiplexer 816 includes first, second and third demultiplexing units (DEMUX units) 810, 812 and 814, respectively, which include the left, center and right field information. Each demultiplexing unit has an input "A" to which the signals YH, YF and YL are applied and an input "B" to which a blanking signal (black signal) BLK is applied. The blanking signal may be a logic O level or ground, for example. The unit 810 extracts the signal YH 'containing the left and right high side field information from the signal YH as long as a signal drive input (SEL) from unit 810 receives a first control signal from a count comparator 817 and thus the presence of left field pixel elements 1 to 84 and right field pixel elements 671 to 754. At other times

causes a second control signal coming from the count comparator 817 that the blanking signal at the input B and not the signal YH at the input A at the output of the unit 810 is applied. The unit 814 and a count comparator 820 similarly cooperate to derive from the signal YL the signal YL 'of the low frequency side panel information. Unit 812 asserts signal YF coming from input A to generate a central field signal YC only when a control signal from count comparator 818 indicates the presence of central field pixels 75-680.

The count comparators 817, 818 and 820 are synchronized with the image signal YF by means of a pulse output signal coming from the counter 8: 2 and responding to a clock signal having four times the chrominance subcarrier frequency (4 xf _sc ) and a horizontal line synchronizing signal H derived from the image signal YF. Each output pulse coming from counter 822 corresponds to a pixel position along a horizontal line. The counter 822 has an initial shift of -100 corresponding to the 100 pixels from the beginning of the negative going horizontal synchronizing pulse at the time T _H s to the end of the horizontal blanking interval, that is, until a time when the pixel 1 appears at the onset of the horizontal line display interval , Thus, counter 822 will have a count of "1" at the onset of the line refresh interval Other counter arrangements may also be developed It should be apparent that the principles applied to demultiplexer 816 also apply to the multiplexer for performing a reverse signal combining operation, as shown in FIG Figure 9 shows an apparatus suitable for performing the NTSC coding process used by the encoder 36 of Figure 1 to generate the C / SL signal place.

In Figure 9, the signals IN and QN appear at four times the chrominance subcarrier frequency (4 χ _sc ) and are applied to the signal inputs of the latches 910 and 912, respectively. The latches 910 and 912 also receive clock signals having a frequency of 4 xf _sc for the transmission (introduction) of the signals IN and QN, and a switching signal having a frequency of 2 xf _sc , which is present at a reverse switching signal input of the latch 910 and at a non-inverse switching signal input from latch 912 is created.

The signal outputs of latches 910 and 912 are combined and applied to the signal inputs of noninverting latch 914 and inverting latch 916 in a single output line where signals I and Q appear alternately. These latches are clocked at a frequency of 4xf _K and receive a switching signal having a chrominance subcarrier frequency of f _K at the inverting or non-inverting input, respectively. Noninverting latch 914 generates an output alternating sequence of positive signals I and Q, and inverting latch 916 generates an output alternating sequence of negative I and Q signals, ie, -I and -Q. The outputs of latches 914 and 916 are combined in a single output line at which an alternating series of paired I and Q signals of opposite pair polarity, i, Q, -I, -Q..., Etc., appear. These signals are used to generate the NTSC coded signal C / SL of the form Y + Y, Y + Q, Y - I, Y - Q, Y + I, Y + Q ... and so on in an adder 918 Luminanzsigr.ai YN combined.

FIG. 12 illustrates a raster imaging apparatus that may be used in conjunction with the time expander and time compressors of FIGS. 6 and 7. In this regard, reference is made to the waveform of Figure 12a, which illustrates the imaging process. In Figure 12a an input signal waveform S is shown with a central portion between the pixels 84 and 670 which is to be imaged by means of a time stretching process at the pixel locations 1 to 754 of an output waveform Y. The endpoint pixels 84 and 670 of the waveform S are mapped directly into the endpoint pixels 1 and 754 of the waveform Y. The intervening pixels can not be mapped directly on a 1: 1 basis due to the time dilation, and in many cases can not be mapped on an integer basis. The last case is illustrated, in which z. For example, the pixel location 85,33 corresponds to the output waveform S of the integer pixel location 3 of the output waveform Y. Thus, the pixel location 85,33 of the signal S comprises an integer part (85) and a fraction DX (0,33), and the pixel location 3 of the waveform Y comprises an integer part (3) and a fraction (O). Figure 12 shows a pixel counter operating at a frequency of 4 χ f _K and providing an output signal M WRITE ADDRESS corresponding to the pixel locations (1 ... 754) on an output screen. The signal M is applied to the Programmable Read Only Memory (Programmable Read Only Memory) (PROM) 1212, which contains a lock table of programmed values that depend on the type of raster image to be performed, such as compaction or stretch. In response to signal M, PROM 1212 provides an output signal as a READ ADDRESS corresponding to an integer and an output DX corresponding to a fraction equal to or greater than zero but less than one. In the case of a 6-bit signal DX (2 ⁶ = 64) is the signal DX for the fractions 0, V (S4,2 / 64, ^3/64 ... ⁶³ / 64th

The programmable read-only memory 1212 permits the stretching or compression of an image input signal S as a function of the stored values of signal N. Thus, in response to the integer values of the pixel location signal M, a programmed value of the signal N READ ADDRESS and a programmed value of the break signal DX are provided. For example, to achieve signal stretching, the programmable memory PROM 1212 is configured to generate a signal N at a frequency lower than signal M. Conversely, PROM 1212 provides signal compression at a frequency higher than that of the signal M, a signal N available. The video input signal S is delayed by the cascaded pixel delay elements 1214a, 1214b and 1214c for generating the image signals S (N + 1), S (N + 1) and S (N), which are mutually delayed versions of the video input signal. These signals are applied to video signal inputs of respective dual access memory 1216a-1216d as known in the art. The signal M is applied to its write address input of the individual memories 1216a to 1216d, and the signal N is applied to a read address input of the individual memories 1216a to 1216d. The signal M determines where the incoming video signal information is written to the memories, and the signal N determines which values are read from the memories. The memories can store one address while at the same time reading out another address. The output signals S (N-1), S (N), S (N + dog S (N + 2) coming from the memories 1216a to 1216d have a time-expanded or time-compressed format depending on the read / write Operation of the memories 1216a-1216d, which depends on how the programmable read-only memory PROM 1212 is programmed.

The signals S (N-1), S (N), S (N + 1) and S (N + 2) coming from the memories 1216a to 1216d are from a four-point linear interpolator, to which the peaking filters 1220 and 1222, a PROM 1225, and a two-point linear in polarizer 1230, the details of which are shown in Figs. 12b and 12c. The peaking filters 1220 and 1222 receive three signals from the group of signals including the signals S (N-1), S (N), S (N + 1) and S (N + 2) as shown, and also receive a peaking signal PX. The value of the peaking signal PX varies between zero and one as a function of the value of signal DX, as shown in FIG. 12d, and is provided by the programmable read only memory PROM 1225 in response to the signal DX. The memory PROM 1225 has a lock table; it is programmed to produce a given value of PX in response to a given value of DX. The peaking filters 1220 and 1222, respectively, provide to the two-point linear interpolator 1230, which also receives the signal DX, the peaked and mutually delayed video signals S '(N) and S' (N + 1). compressed or stretched) video output signal Y, wherein the output signal Y by the expression

Y = S '(N) + DX [S' (N + 1) -S '(N)]

is defined.

The described four-point interpolator and peaking function advantageously approximates a (sin X) / X 'interpolation function with good resolution of the high-frequency detail.

Figure 12b shows details of the peaking filters 1220 and 1222 and the interpolator 1230. In Figure 1220, the signals S (N-1), S (N) and S (N + Dan) are applied to an evaluation circuit 1240 where these signals pass through the peaking coefficients - ¹ A, V / 2 and ^-1 , respectively, As shown in Figure 12c, the evaluation circuit 1240 comprises the multipliers 1241a to 1241c for multiplying the signals S (N-1), S (N) and S (N + 1) by the ¹ A, ¹ AbZW.- ¹ A. The output signals coming from the multipliers 1241a to 1241c are added in an adder 1242 for the generation of a peaked signal P (N) The signal P (N) is signaled by the signal 12B for multiplying a "peaked" signal, which is added to a signal S (N), to produce a peaked signal S '(N) is operated in a similar manner, in the two-point interpolator 123 0, the signals S '(N) of S' (N + 1) in a subtraction circuit 1232 for generating a differential signal, the output signal coming to the signal DX in a multiplier circuit 1234 in the adder 1236 to generate the output signal Y to the signal S ' (N) added.

In the following, reference is made to FIG. A broadcast compatible wide-screen television signal is received by an antenna 1310 and applied to an antenna input of an NTSC receiver 1312. The receiver 1312 normally processes the compatible widescreen signal to effect 4: 3 aspect ratio image reproduction, with the widescreen side field information partially crowding (ie, the low frequency information) in the horizontal oversampling areas that are out of view of the observers and partially (ie, the high frequency information) are included in the modulated alternating subcarrier signal which does not disturb the standard receiver operation.

The compatible widescreen signal received by antenna 1310 is also applied to widescreen receiver 1320 capable of reproducing a video image having a large aspect ratio of, for example, 5: 3. The received wide image signal is processed by an input unit 1322 including a high frequency tuner (RF tuner) and an amplifier circuit and a video demodulator producing a baseband video signal. The baseband video signal coming from unit 1322 is converted to the digital (binary) using an analog-to-digital converter (ADC) 1324, which operates at a sampling frequency equal to four times the chrominance subcarrier frequency (4 xf, _c ). Shape converted. '. A widescreen digital video signal coming from the output of analog-to-digital converter 1324 is applied to HVT bandpass filter 1326 for high side field signal (SH) filtering. The filter 1326 has the construction according to FIG. 10c and a passband of 2.368 MHz ± 70 kHz. The high frequency side panel signal is applied to one input of a subtractive combiner 1328 whose other input from the analog to digital converter 1324 receives the widescreen video output signal after being delayed by a signal processing delay compensation unit 1330 caused by filter 1326.

The NTSC format output C / SL coming from the combiner 1328 contains the full bandwidth central field information as well as the low frequency compressed side field information. The signal C / SL corresponds to the signal C / SL coming from the output of the encoder 36 according to FIG. In the remainder of the discussion of Figure 13, the signals having counterparts to the signals of Figure 1 are designated by the same characteristics. The signal C / SL is converted to the subcarrier signal SC by means of a decoder circuit comprising an HVT bandpass filter 1332 having the construction of Figure 10c as well as a passband of 3.58 + 0.5MHz, subtractive combiner 1334 and synchronous quadrature demodulator 1336 addresses, decodes into its constituents YN, IN and QN. Filter 1332 separates the chrominance component from the signal C / SL. The luminance signal YN is obtained by subtracting the separated chrominance output from the filter 1332 from the signal C / SL in the subtractive combiner 1334 after delaying the signal C / SL from the network 1333 to equalize the propagation delays of the signals applied to the combiner 1334.

The separate chrominance signal from filter 1332 is demodulated by demodulator 1336 to produce color difference signal components IN and QN. The signals YN, IN, and QN are separated into compressed low-side field information YO, IO, and QO and stretched central field signals YE, IE, and QE by means of a side field central field signal separator (time divisional multiplexer) 1340. The demultiplexer 1340 may be operated based on the principles of the demultiplexer 816 of FIG. 8 previously discussed.

The signals YO, IO, and QO are time stretched by a time expansion factor 1342 (corresponding to the side-compression factor of the encoder of FIG. 1) to reflect the original spatial relationship of the low-frequency side-field information in the wide-screen signal as represented by the restored low-frequency side-field signals YL. IL and QL is presented to restore. Similarly, the YE, IE, and QE signals are time compressed with a central field compression factor (corresponding to the center field expansion factor in the encoder of FIG. 1) by time compressor 1344 to provide space for the side panels, around the original spatial

Relationship of the central field signal in the wide-screen signal, as represented by the restored central field signals YC, IC and QC restore. Time compressor 1344 and time expander 1342 may be of the type shown in FIG. The restored low frequency side panel information YL, IL and QL are combined in a combiner 1346 with the restored high frequency side panel information YH, IH and QH obtained as follows. The high-frequency side-field signal SH coming from the filter 1326 is amplified by an amplifier 1350 to compensate for the attenuation with a gain 4 produced by the attenuator 44 in the encoder according to FIG. The amplified high frequency side field signal is quadrature demodulated by a demodulator 1352 and then corresponds to the alternate subcarrier signal ASC. The demodulated high frequency side field signals X and Z are applied to a luminance chrominance separator 1354, such as a demultiplexer of the type shown in Figure 8, for the generation of high frequency side field luminance components LH and RH and color signal components IH and QH.

Specifically, the signal X for the generation of signals LH and IH is demultiplexed and the signal Z is demultiplexed for the generation of signals RH and QH, as illustrated in FIG.

The left and right high frequency side panel signals LH and RH are time compressed using a time compressor 1356 having a page compression factor (corresponding to the page expansion factor in the encoder of FIG. 1). The compressor is of the type discussed in connection with Figure 12 and maps the right and left compressed high frequency side panel signals in place in the respective horizontal scan lines, thereby producing the spatially restored high frequency side panel signal YH.

The spatially restored high frequency side panel information YH, IH and QH are combined with spatially restored low frequency side panel information YL, IL and QL to generate reconstructed side panel signals YS, IS and QS using combiner 1346. These signals are spliced using a splicer 1360 to form a fully reconstructed wide-field luminance signal YF and fully reconstructed wide-field color difference signals IF and QF. As will be seen from the following discussion of the splicer 1360 of Figure 14, the splicing of the sidebase and central field signal components is performed in a manner that actually eliminates a visible edge at the boundary between the central field and the side fields.

The wide-screen signals YF, IF and QF are converted to an analog form by means of a digital-to-analog converter 1362 before being applied to a video signal processor and matrix amplifier unit 1364. The video signal processor component of unit 1364 includes the signal gain, DC level shift, peaking, brightness control, contrast control, and other conventional video signal processing circuits. Matrix amplifier 1364 combines the luminance signal YF with the video signal R, G, and B to produce the color image processor component Color difference signals IF and QF. These color signals are amplified by the playback driver amplifiers in unit 1364 to a level suitable for directly operating a wide-screen color image display device 1370, such as a widescreen display tube.

In Figure 14, splicer 1360 is shown connected to network 1410 for producing the full bandwidth luminance signal YF from sidewall luminance signal component and central field luminance signal component YC, and I signal splicer 1420 and Q signal splitter 1430, which are similar in structure and operation to network 1410 , As previously mentioned, the central field and the page fields are intentionally expanded by several pixels, e.g. B. 10 pixels, overlaps. Thus, the central field and side panel signals have multiple redundant pixels prior to splicing during the signal encoding and transmission process.

In the wide-screen receiver, the central field and side panels are reconstructed from their respective signals, but due to the temporal stretching, temporal compression, and filtering performed on the field signals, multiple pixels at the side panel central panel boundaries are corrupted or distorted. The overlap areas (OL) and the corrupted pixels (CP) (somewhat exaggerated for clarity) are indicated by the waveform associated with the signals YS and YC in FIG. If the fields did not have an overlap area, the spoiled pixels would be contiguous and a seam would become visible. It has been found that an overlap area of ten pixels wide is large enough to compensate for three to five corrupted boundary pixels.

The redundant pixels allow the merging of side panels and central panel in the overlap area. A multiplier 1411 multiplies the side field signal YS by a weighting function W in the overlap areas, as illustrated by the associated waveform, and before the signal YS is applied to a signal combiner 1415. Similarly, the central field signal YC is multiplied by a multiplying unit 1412 with a complementary evaluation function (1-W) in the overlap areas, as illustrated by the associated waveform, and before the YC signal is applied to the combiner 1415. These evaluation functions have a linear ramped characteristic across the overlap areas and contain values between 0 and 1. After evaluation, the page and central field pixels are summed by the combiner 1415 so that each reconstructed pixel is a linear combination of the page and central field pixels.

The evaluation functions should preferably approach one near the innermost boundary of the overlap area, and approach zero at the outermost boundary. This ensures that the corrupted pixels have a relatively small effect on the reconstructed field boundary. The illustrated linear ramped function fulfills this requirement. However, the evaluation functions need not be linear, and a non-linear evaluation function with curvilinear or rounded end sections, i. with sections near the evaluation points 1 and 0 can also be used. Such a weighting function can be readily accomplished by filtering a linear ramped weighting function of the kind illustrated.

The evaluation functions Wound 1-W can be readily generated by a network including a lock table and a subtractive combiner, the network responding to an input signal representing the pixel positions. The page field central field pixel overlaps are known, and the lock table becomes so

is programmed to provide output values from 0 to 1 corresponding to the weighting function W in response to the input signal. The input signal can be developed in a very different way, for example from a counter which is synchronized by the individual horizontal line synchronizing pulses. The complementary evaluation function 1-W can be generated from the subtracted evaluation function W from one. The principles of the disclosed invention are applicable to other types of standard broadcast television systems, such as PAL.

Claims (4)

An arrangement for processing a widescreen video signal from which a Breitbüd with an aspect ratio greater than the standard aspect ratio, arises, the widescreen video signal having a central field signal component with a given aspect ratio and a side field signal component with low-frequency and high-frequency components, using the coded from each other different processes, characterized by means (1322 to 1342) for decoding the low-frequency side field signal component; means (1350 to 1356) for decoding the high frequency side panel signal component; means (1344) for modifying the aspect ratio, the central field signal component; and means (1346 and 1360) for combining decoded low and high frequency side field signal components with the central field signal component with the modified aspect ratio. V 2. Arrangement according to claim 1, characterized in that the given aspect ratio of the central field signal component is a standard aspect ratio of approximately 4: 3; and in that the modifier (1344) times the aspect ratio of the central field component. 3. Arrangement according to claim 1, characterized in that the low-frequency component (YL) is coded by the temporal compression of the low-frequency component in the horizontal Bildabtabtastgebiet. 4. Arrangement according to claim 1, characterized in that the high-frequency component (X, Z) of the side field signal component is coded by modulating an alternating subcarrier (ASC) with the high-frequency side field component, wherein the alternating subcarrier has a frequency which differs from the chrominance Subcarrier frequency is different.