JPS63502156A - Bandwidth compression for television signals - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] This invention relates to bandwidth compression for television video signals. If HDTV is to become a practical product, it must be transmitted to homes. Some way to do so is required. Such transmissions are carried out where extra radio frequency range is available. Since this is not available either, it is generally assumed that the method will be via satellite. this place The main problem is the extremely wide bandwidth of the original HDTV signal - possibly the combined brightness. and about 40 to 50 MHz for color (chrominance) information. this Such a bandwidth can be accommodated in the 12 GHz DBS (Digital Satellite Broadcasting) band. I can't do it.

The high transmission range required for HDTV is limited only to terrestrial and satellite broadcasting. Signals via other media such as videotapes, videodiscs and cables It also causes problems for spraying. In order to overcome these difficulties, there is a certain form bandwidth reduction is required.

First-order, second-order and third-dimensional pre-filtering together with VC+ 5ub-NyquiSt) Describes how to reduce bandwidth using sampling. Ru. Mobile adaptive prefiltering techniques have been described relatively recently, e.g. NHK Research Institute Note No. 304, 1984, “Single Channel” by Ninomya et al. Channel HDTV broadcasting method J (NHK Laboratory Note A3 04゜1984, NINOMIYA et al” A single C hannelHD TV Broadcast System”) and 198 June 5th MO (MU SE) J (& paPgrpresented by FUJIO.

SUGIMOTOana NINOMIYA at, the l 4-eh: Engineering international Te1elViSiOn symposium, M OntreuX.

June 985, “HDTV’rranmission meth oa(MUSE)''). This method is considered to be of little use to the eye. It is based on the filtering removal of image frequency components. filter signals have much reduced bandwidth and require relatively low speeds for transmission. can be resampled at degrees.

The use of mobile adaptive spatial subsampling provides an impressive reduction in transmission bandwidth. Be able to do it. Although static areas of the image are transmitted at full resolution, this information can be transmitted in many different ways. distributed in the television field. The moving area is the area in the moving object. Transmission at reduced resolution takes advantage of the reported inability of the eye to perceive details. sent. This was used in NHK's MUSE method (described in the paper above) The strategy has a target bandwidth compression ratio of 4:1 and a bandwidth of only 8.1 MHz. Becomes the final transmission bandwidth. This method is generally promising, but variable. Give the result that something is possible and unpredictable. As evaluated by the machine The performance of this method varies depending on the receiver, and this means that this method is Which areas of the scene are moving and which areas are stationary at the receiver of This occurs in order to detect. Furthermore, this method Since it requires a movement detection circuit, it is relatively expensive.

(b) The invention is defined in the following claims, which should be referred to here. be.

In selected embodiments of the invention, described in more detail below, the bandwidth is reduced. The analog video signal to be processed undergoes two types of filtering in parallel: Temporal filtering, which is appropriate when static parts are involved, and image movement Subject to spatial filtering, which is appropriate if the parts are involved. any Which filter output is used at a moment depends on the image content and The filtering control signal is used not only in the transmit km signal but also in analog signals. The digital signal is transmitted to the receiver/'3 as a digital signal associated with the video transmission signal.

The method adopted is that highly detailed images are transmitted in static and slow moving areas of the image. make it possible to The remaining areas are details that emerge, shot changes and unusual movements. Contains information with poor correlation, such as This area is where the image details are in the field. The human visual system really does not require much detail when there is poor correlation between the two objects. Therefore, it is transmitted in relatively low definition.

The reduced bandwidth signal consists primarily of analog subsampled image data; It also includes digital control signals that have proven to be critical to the success of the seed scheme. I'm here. The digital signal is transmitted using a coding method for each moving area. Conveys reliable information about the moving content of images and details.

Since the encoder and decoder can work in harmony with each other, the receiver/decoder The correct post-filter in can be chosen for each area to display the bandwidth-reduced signal. Restore the format. This predicts movement from the video signal received at the receiver. The need to try? However, the received signal has no information available at the encoder. Since it contains only one-fourth of the sound level and also includes the acceptable Kl sound level, it can be transferred reliably. It is difficult to predict the movement. Also, the transmitted signal is subsampled by a thermally unknown method. ing. By transmitting a digital signal associated with an analog video signal , all decisions about the moving content of the scene are made in the encoder. The receiver is Just be told how to decode the signal. With this 5KI, the receiver has been simplified. At the same time, the overall system performance is improved by reliable tracking between encoder and decoder. be good. Additionally, display improvement techniques such as line and field rate up conversion , as reliable movement information that was previously missing is readily available from the data channel. , simplified.

Brief description of the drawing The invention will be explained by way of example with reference to the drawings, in which: FIG. 1 is a circuit diagram of a transmitter/encoder embodying the present invention. FIG. 2 shows a receiver embodying the invention that can be used with the transmitter/encoder of FIG. Fig. 3 is a circuit diagram of the device/decoder; (a) by multiplexing the digital information signal into a video signal and FIG. 4 is a circuit diagram illustrating two methods of transmitting data from the The system according to the invention uses an interlaced (interlaced) signal that passes through the transmitter. It illustrates possible upstream routes from the mountain.

The illustrated television system includes an encoder and an encoder shown at the transmitter in FIG. Each receiver is provided with a decoder as shown in FIG.

The signal from the HDTV source 10 is then digitalized for ease of processing in the ADCII. is converted to digital form and first applied to a diagonal spatial filter 12 to filter the signal. Diagonal components in the spatial content are reduced. This results in an initial 2:1 bandwidth reduction. Small is given. After this diagonal filtering, the signal is divided into two two-dimensional prefilters. 14.16vc is added. The first @-position filter 14 is a vertical one-time filter. The purpose of this is to reduce the signal bandwidth for stationary and slow moving areas. and the second prefilter 16 is a vertical-horizontal filter or a spatial filter. This reduces the signal bandwidth for faster moving areas.

The two filters 14.16 may actually include certain elements in common, and Some parts of the circuit configuration can be shared. Includes appropriate compensating delay 15 .

These filters are from BBC Research Department Report No. 1975/36 [PAL color one type ``Filtering of luminance and color signals to avoid cross-color in Written by Dr. H. O. Dreweri, published in 1975 (BB CRe5earch 　DepartmentRepOrtA　1975　/36　”　the　fi ltXing of luminance and c button rominance ignal to avoid cross-colourin a P A L color system”, Dr. J. O. Drewery.

It may be based on the filter described in published 1975). .

The reduced bandwidth signal is resampled at a relatively low rate in a resampler 18.20. to form two alternative transmission signals.

Resampling is a relatively coarse-grained image that is sufficient to support reduced bandwidth after filtering. It is used as a lattice.

The outputs of the two resamplers 18 and 20 are then combined to form the video component of the transmitted signal. It is added to a combiner 21 and a channel encoder circuit 22. For now, the circuit 21 is Whether that part of the signal represents stationary or moving information, respectively. resampler 18 or resampler 20 depending on the control input 24 indicating It can be thought of as a selector switch.

The resulting video is converted to analog format at KDA023 for transmission. a digital signal that is converted and indicative of the value of the control signal 24 at the channel encoder 22; be increased by

In the receiver, the received broadcast signal is sent to the receiving input after the normal RF and IF stages. Separate the analog and digital components of the force signal. Digital components are digital It is applied to a control block generator 32 which reshapes the control signal 24. analog component is converted into digital form by ADO 31 and then added to interpolators 34 and 36. However, this interpolator has two advantages: 1. The sampling signal is also directly transmitted to the receiver. A displayable signal is reconstructed from the received signal using σ, which is suitable for close display. interpolation Interpolator 34 is a temporal interpolator for use in static image areas, and interpolator 36 is a temporal interpolator for use in static image areas; is a spatial interpolator for use in areas. The interpolator receives input from source 10 Return the resampled signal to.

The recombinant circuit 38 is controlled by the control block generator 32, but for the time being selects the output of the interpolator 34 or the output of the interpolator 36 and receives it via the DAC 39. It can be considered as an addition to the machine display device 40. The recombination circuit 38 Since it responds to the same control input 24 for digital information transmitted as a It operates in exact synchronization with the combiner 22 in the encoder.

The extraction of control signal 24 will now be described. This is the interpolator 34 in the decoder.　 , 36 and the same interpolator 44.46 can be used to represent the resampled signal in the encoder. In principle, this can be achieved by restoring the original format. Therefore, the encoder Use the signal that would be generated by the decoder if a proper type of interpolator were used. Can be done. The next pressure subtractor 48 subtracts the input signal from the output of the interpolator 44, and The input signal is subtracted from the output of the interpolator 46. Output of subtractor 48.50 , two respective error measuring circuits 52,54 measure the output and input signals of the associated interpolators. Determine the measure of "error" by the mean square or mean modulus between the sand That is, circuits 52 and 54 are connected to the output of interpolators 34 and 36 in the decoder as much as possible. We believe that this signal (assuming a thermally perfect transmission channel) represents the signal that is appearing. It can be compared with the original source signal. The error processor or comparator 56 Which of the errors is more acceptable in response to relatively greater fidelity in the resampled signal? Depending on this determination, a control signal 24 is generated.

The encoder has two monitor displays that can view the two interpolated signals. 58,60 can be provided, but these are also limited to g18 minμ for circuit operation. Not useful.

In this all-purpose solution for HDTV, all encoding decisions are 11 primitive undistorted. This is done at the encoder where the signals are available for comparison. The sending link is high or low is formed as an analog image signal augmented by a certain digital control signal. We have applied this method to digitally assisted television (Digi tall). It is called yASSiSted Te1evlsion) (D A T V) I'm here. The decoder has much reduced information content and much poorer signal-to-noise Based on the transmitted signal with the sound ratio, the decision itself can be displayed instead of requiring 7 things to be done. simply be told how to reconstruct the possible signal.

In this way, system performance is increased while reducing receiver cost and complexity. Be reduced.

The system illustrated in Figures 1 and 2 is thus a hybrid analog/digital code. It can be seen that it is an encoding method. The moving area of the image is transferred to the bandwidth compression encoder in Figure 1. Stay and be worked out. The subsampled video is sent in analog format, and there is no movement The signal from the encoder indicating whether the area is considered to be be sent as a digital signal. Analyzer of the integrated transmission standard forming two parts of the hybrid The combination of the log part and the associated digital part, the digital moving signal, has a considerable capacity. used to regenerate the original HDTV standard at the receiver. Let's go. The receiver is simplified as it no longer requires a movement detector, and it Make their behavior more predictable. At the same time, this technique achieves better final results However, this increases the amount of moving detectors available before subsampling. Not only does it work more satisfactorily when given the amount of information that the encoder This means that more complex, more expensive, but higher performance algorithms have been installed. This is to enable the use of As shown, the two modes of reoccurrence of the scene, In other words, the stationary mode and the moving mode are uncompressed in the encoder. Area by area is compared with the original signal. The stationary mode has a relatively small mean squared error. If it is, then that particular area is considered stationary, and vice versa. The same is true.

The prefilter 14 operates effectively in a static area, e.g. Spread the transmission about the four fields in a quincunx pattern. That's it A complete, highly detailed still image can be stored over the same number of fields and over a period of time.

In moving areas, priority must be given to smoothing the depiction of motion. No. Bandwidth reduction is therefore achieved by the pre-filter 16, resulting in an increase in spatial resolution. This must be achieved by a reduction.

The prefilter, subsampler and interpolator can be used, for example, in the N)I K　It can be similar to the one used in the MUSE method, so please explain it in more detail. is not preached.

Encoding decisions are made in a block-by-block manner for each block of a certain number of pixels. Let's go. The size of the block may vary when transitioning from one transmission mode to another. Prefilter and interpolator apertures to minimize the appearance of Kero artifacts VCf should be similar to It shouldn't be so big that it hurts. Vertical and horizontal dimensions approximately 1/100 of the image height The method is a good compromise and provides a control chaffle with a Give a range of manageable data rates.

It is preferable that the blocks are diamond-shaped Vcf squares, but this is because this shape is the diagonal line between the blocks. This is because it tends to make the connection part more difficult to see. Therefore, 32 pixels in total A diamond-shaped block of 8×8 pixels was adopted, containing approximately 55,000 pixels per image. 7 ri will be given to you. In any case, at least 100o blocks per image. and each block contains more than 10 elements, preferably more than 25 elements. desirable.

Combiner 22.38 is simple to avoid sudden transitions that may be noticeable. It can operate as a cross filter rather than a selector switch.

Sources and displays can operate to any of a range of desired high-definition standards However, these do not have to be the same. The source standard was used in HDTV studios Depends on the standard. The source standard is 1125/60/2:1, i.e. 1125 One image per book, 60 fields per second, 2 interlaced fields for each image. Assuming 1035 valid lines, 30 MHz horizontal bandwidth (4 degrees) and 26 There will be a valid line connection time of μs. According to sampling theory, this is 1 Sampled without distortion at a rate of 560 samples/active line, i.e. 60MH2 Satl is enough. In practice some additional margin will be required.

For windings where this is about 10%, the number of specimens per effective line rises to about 1700.

Alternate transmission is assumed for the two chrominance channels, and each chrominance signal is a luminance signal. If you have half the horizontal resolution of The information will be 2550 samples/active line. When line and field blanking is omitted , the total video sampling rate is 1035x30x2550, or 79.2x1 0 samples per second. This is coded as an analog signal without subsampling bandwidth compression. If this were to occur, a bandwidth of just 40 MHz would be required.

The required 4=1 bandwidth compression can be achieved in two stages. 2:1 first is achieved by diagonal filtering in filter 12. 2:1 The second rate of This is achieved by more severe diagonal filtering operating on moving parts. this After the filtering process, the signal is subsampled and then reduced to a bandwidth of only 10 MHz. can be transmitted in analog format.

This allows bandwidth up to 12MHz to be utilized without causing excessive levels of interference. This corresponds to approximately 80% of the capacity of a 27 MHz wide WARC channel. do.

1 Mbit/s capacity, with extenuating synchronization code, clamp period, etc., for dual channel If channel digital audio is used, the VC will have approximately 2 Mbit/s of auxiliary mobile data. should be available for data. This is true for incoming HDTV signals. For example, a diamond-shaped block of 4 fields x 8 pixels is encoded with a 1-focus signal. indicates whether the information in that block is moving or stationary. 1035x1700x30/32f or 1.65 megabits /sec) and still allows some spare capacity to remain.

Multi-tap if Vc1920 specimen/active line is used for studio standard It would be possible to transform the 1700 samples using interpolation, but another method The bandwidth compression is based on 1920 samples/effective lines, and the chrominance The number of samples is reduced to one third of the brightness rate, i.e. 640 obscurations/active lines. You will be able to do that. The overall video bandwidth is hardly affected in this case. Sending The selection of the higher frequency sampling station wave number before the signal increases the visibility in the diagonal filtering stage. It will help reduce information loss.

Specimens used in source and display devices are not limited by this invention to any temperature range. Can be made to suitable higher definition standards. In particular, the standard is It can be standard (interlaced) or sequential (non-interlaced).

Turning now to FIG. 3, two ways in which hybrid encoding can be implemented are shown.

The first method shown is the one used in the system of Figures 1 and 2. be. An incoming HDTV signal is detected and an appropriate bandwidth compression method is used to select it. added to the circuit 100 for measuring parameters in image content, This circuit is effectively a movement detector in this case. The HDTV signal is then e.g. Receives essentially analog processing in processor 102, including sequential-to-interlaced conversion. and is added to multiplexer 104. In reality, this process is a digital technology. It can be executed using The output of the parameter measurement circuit 100 can be any desired Digital processing 1 For example, the process of relating parameters to corresponding areas of the displayed image. block format, and then applied to another input of multiplexer 104, where combined with an analog signal for transmission over a single channel. time division or Any conventional method may be used, such as frequency division multiplexing. [shown in al. The method described above is the most useful for broadcast purposes.

However, multiplexer 104 is omitted as shown in FIG. It is also possible to transmit digital data signals by separate channels. this is It may be more appropriate in situations such as videotape recording. for example , the analog signal may be a standard interlaced signal recorded in the usual way, and the data channel to a convenient audio channel or to the first theme by an appropriate time code. can even be recorded onto another tape, locked to the tape.

Although this specification refers to "transmitting" a signal for simplicity, However, it is recognized that the term includes other forms of processing such as recording signals. It will be.

Information about which parts of the incoming image are moving and which parts are stationary to inform the receiver decoder how to correctly decode the bandwidth compressed HDTV signal. is transmitted in the digital part of the mixed signal. However, it is not received It can also be used for stretching in machines.

Similarly, when high-frequency static detailing places heavy bandwidth demands on analog channels, The nature and location of these features are then transmitted over a digital channel for final reconstruction. It can be used when constructing.

A feature associated with sequential scanning is that the transmission is interlaced to help conserve bandwidth. - If the up-converted sequential representation is If you transmit it using a regular channel ().

The above-mentioned hybrid method will become a progressive method for improving television as technology advances. It is also effective even if In other words, the logical progression of the HDTV standard is shown in Figure 4. It can be predicted that Initially, the source, transmission and display are all the same (shown in row 1 of the figure). (interlacing) standard. However, the data channel Some advanced plans include the progressive introduction of more advanced methods capable of higher resolution. It will open the way for At the same time, possibly with a relatively small display device. The second receiver may not need the full resolution and therefore all incoming signals This makes it much easier to control the decoder to handle it as if it were moving. It may be possible to use a data channel to

As receiver technology advances, the situation progresses to that shown in row 2, and the receiver becomes more reliable. The signals should be sequentially up-converted to constant distortion for display. Transmission field rate 5Q) (If lower than Z, the It may also be necessary to increase it.

Further advances will allow the studio signal processing chain to operate in a sequential fixed-course fashion. That would be the case. Two options are shown in rows 3a and 3bVc of the diagram. camera is scanned in j-sequential fashion, and the result is converted to interlaced fashion, or An interlaced signal is generated from an interlaced source, possibly using complex motion estimation methods.

In the former case, the auxiliary data signal is carried along with the interlaced video. This is a simple addition. This data signal facilitates subsequent reconversion to progressive scan format. -f and is conceptually the same as mobile data carried by a transmission system. be.

In fact, a bandwidth compression encoder for a transmission line is generated using a sequential-interlaced conversion scheme. It would work better if it were allowed to use the data. This information The flow is shown by the dotted line in FIG.

The final advancement would be to use sequential formulas throughout the studio. This is shown in the bottom row of the diagram. has been done. In this case, the information in the digital channel is based on the large amplitude of the original signal. Primarily aimed at reducing the bandwidth requirements associated with dynamic information and high frequency stationary information. It will be related.

The method for changing the display of species is as described in the BBC Research Department Report No. 198378. ・Improved display device J for r625 television images written by Roberts (BBCRe5 search Department Report Al 983/8, “Th e　JmprOVed　Displayof　525-1inetelevis ion PiCtureS' by A, Roberts) VC listed ing. Some of the relevant information to aid the receiver amplifier conversion is the analog signal itself. While the hybrid transmission can be extracted from the body, the hybrid transmission Allow additional substantive information to be transmitted. Furthermore, the bandwidth provided by mixed transmission There are several system-wide advantages unrelated to compression. For example, motion detection The responsibility for this is transferred to the signal encoder. only a few of these are needed This can significantly increase cost and complexity. Therefore, and More complex movement detectors with better performance can be used depending on the .

The scheme outlined above and illustrated in Figures 1 and 2 has two potential drawbacks. ing. The first of these is that the signal remains even when the image is completely still. It is subject to losses associated with stationary diagonal filters. This is this The filter should also be motion-adaptive, and in fact a diagonal filter for very slow movements. This can be overcome by a Kj filter so that the data is completely removed. Additional travel information may be added to the auxiliary data channel. Additional bandwidth available To achieve this, either slightly increase the compression ratio of the subsampled video or It may be necessary to increase the size of the individual blocks into which the image is divided. but , use the same block dimensions for the additional movement signal as for the main movement signal. There is no need to For example, the main movement signal uses a block of 4 field lines x 8 pixels. and additional movement signals for low speed are used in blocks of 8 lines x 16 pixels. be able to. This is the extra capacity that must be accommodated by a factor of 4. (0,42 Mbit/s VC).

The second drawback of the scheme outlined above is potentially more serious.

In the past, the eye was able to take advantage of high spatial frequencies when they were moving. It is generally believed that this is not possible, and this is actually true for fixed fixation points. Ru. However, for ordinary television images, the eye is exposed to general VC, low speed For those interested in continuous motion or for high velocities with "intermittent" motion. attempts to track moving areas. That is, when the eyes are tracking a moving object, , the image on the retina is static and loss of fine detail can become noticeable. Ru. The spatial detail discrimination requirements of the eye in moving areas have not been assumed in most of the literature. is sufficiently likely to be reduced by a large factor that is It is certainly not reduced in the case of relevant movements.

This problem also comes at the cost of requiring increased capacity in the auxiliary data channel. can be overcome by paying. In this case, the amount of movement that a given area of the image receives is The motion representing the amount of interdigitation can be estimated to generate the vector. October 1986 Presented at the 128th SMPTE Technical Conference, New York, March 24-29. paper, reprint number 128-49 [Adaptive Subsampling and Motion Compensation DATV "HDTV Bandwidth Reduction through Technology" (a paper presented at the 12B th SMPTE Technical Conference e, 0ctover 24-2 g, 1 g F36 New York, 　Reprint　j=　128　49　,　”HDTV BandWidt hReduCliOn by AdapjiVe Subsampluig a nd Motion compensation D A T V TeChni queS” and my UK patent application 86173201C describes a method. Ru. In this method, two consecutive images are correlated as a function of displacement and One or more peak correlation values corresponding to each motion vector are determined. The image is Which of the possible vectors is then analyzed in a block of few pixel squares? It is determined which is most appropriate for the block. Other methods that can estimate motion vectors The method was introduced by Sugi at the 14th International Television Debate in Montreux in June 1985. The paper “Second Generation HDTV Standard Conversion” submitted by Moto, Fujio and Ninomya Vessel J (a paper presentato the 14th In ternatiOnal Te1evision Sympo-sium, M Ontreux, June 1985 t)y SUG IMOTO.

FUJ I Oand N lN0M IYA,”5econd-ge nerationHDTV 5standards converter”) known from.

Such techniques can be used, for example, to identify the two most important moving items in a scene; One of these is the motion base of a moving item, which could be the background during a camera pan, for example. It can be used to measure vectors. These vectors represent each field It only needs to be sent to the receiver once each time, requiring very little additional data capacity. do not have. If each image block is then coded with 2 bits instead of 1 For example, there are four different states - ``at rest'', ``moving with vector 1'', [vector "Moving with vector 2", and "moving with vector 1 or 2" - be able to. The address supplied in the field is stored in the transmitter encoder. and the receiver decoder is detected as moving with vector velocity 1 or 2. Those areas are encoded using a stationary algorithm, so the system is Adjust to allow reproduction at some full still resolution.

Strictly speaking, this strategy requires that the speed of motion is constant with an integer number of pixels per field. work only for However, using interpolation, fractional pixel components per field are He can handle the speed he has. However, the main component of scene motion is The removal marcoder aids in the selection of diagonal filtering parameters, and In any case, the essential effect of the subpixel movement error is sufficiently favorable to eliminate the unnecessary Kj of this interpolation. It would be convenient.

The way this method uses motion vectors means that the moving area appears stationary. The reference frame of the high spatial detail rendering filter 16 can be eliminated in the filtering method. Transfer The motion area may then be transmitted in full spatial detail at its correct spatial location.

The decoder knows the value of the vector for the next field of motion during the vertical blanking period. and which area belongs to which interlock x vector via the control data channel. Continuously informed. It then converts the complementary excluder to the reference frame of its stationary interpolator. Reconstruct highly detailed moving areas in addition to the moving area. prefix field for exclusion data and interpolator reference frames are used instead of the bandwidth-compressed signal to receive motion-compensated signals. This allows the machine to use motion-compensated signals, thereby preserving compatibility. Ru. In such a reception, the motion-compensated area is used without changing the motion regime. decoded so that these areas are not in the correct position but in the previous description. This allows the images to be reproduced with relatively poor spatial resolution.

To determine whether motion is present and what type of motion is present In fact, the sizes of the blocks used do not need to be the same.

It should be noted that both of these image quality improvements require sufficient data capacity. If VC is assigned at the start of any broadcasting operation, this can be achieved in a compatible manner. can be done. A receiver that is not equipped to process additional slow-moving information may simply It is obvious to continue filtering all still images diagonally.

In a similar manner, receivers that are not equipped with motion vector information can use All areas of the image that have not been identified as 'static' are simply is also decoded as if it were moving in an unmeasurable vector. The paraphrase f is “ "Moving with vector 1" and "Moving with vector 2" are interpreted as "moving" in 2018. Be it. This principle is, of course, relatively useful for transmitting scenes involving complex movements. It could be extended to include multiple motion vectors.

The arguments presented so far have been directed towards the transmission of high-definition images (although This digital support principle is similar to MAC or current PAL and NTSC. It can be used to improve the performance of systems with lower basic resolution. You can do it. After digital codification, simple mobile signals sent through channels are Amplifier conversion of displays to high line counts or progressive scans with motion and high frequency vertical detail It could be made more effective by helping to distinguish between as a result The potential increase in vertical detail that can be obtained can be increased horizontally by using bandwidth compression techniques. direction, but here again a reliable moving signal is required. It must be heavy.

For example, C-MAG blocks motion vectors with move/standstill instructions. It can be sent in a method for each message. Teletext transmission via PAL be included in the motion vector information, e.g. for improved PAL decoding or up A receiver motion detector may be used to assist in the conversion.

Further addition of motion vector information enables accurate processing of correctly positioned moving objects with sufficient processing power. It is possible to interpolate intermediate fields with Vcf. This means that adjacent moving frames caused by simple, motion-corrected, temporal interpolation between fields. Removes large artifacts such as combing of moving horizontal details. There will be.

A reliable motion signal also eliminates the complex noise measurement circuits required for remote motion detectors. Easily achieve video noise reduction at the receiver by avoiding the need for It would be possible to convert it into

folding screen package international search report AN) JEX τO’=ε INTERNATIONAL SE, “1RCF, RE? ORτON

Claims (18)

[Claims] 1. At the transmitter, video signal source, The bandwidth of the video signal generated by the source is divided into two or more different bandwidths in response to a control signal. a bandwidth compression device for selectively compressing the bandwidth by a method according to the invention; device for transmitting bandwidth compressed video signals in analog form; A bandwidth compressor is configured to transmit bandwidth to a bandwidth compressor in response to the image content of a video signal generated by a video signal source. a control signal that generates a control signal commanding the selection of the desired one of the width compression methods; It is equipped with a generator, And in the receiver, Apparatus for receiving transmitted bandwidth-compressed video signals, bandwidth-reduced transmitted video The signal can be selectively displayed in two or more different ways depending on the image content. a device for reconstructing a signal for comprising an apparatus for displaying the reconstructed signal, and further comprising: transmitting a digital signal including said control signal in conjunction with an analog video signal; equipment at the transmitter for the output and select the desired one of the reconstruction methods to the reconstruction device and reconstruct the signal for display. a video signal transmission system comprising: a device in a receiver that directs the configuration; . 2. A bandwidth compressor is installed in a vertical line, both connected to receive signals from the sources. The method according to claim 1, comprising a temporal filter and a spatial filter. 3. with a diagonal filter between the source and the vertical temporal filter and spatial filter The method according to claim 2. 4. The diagonal filter adaptively depends on the movement and the data transmitted to the receiver. Claims wherein the digital signal contains information about the mode of operation of the diagonal filter. The method described in section 3. 5. Claim 1, wherein the control signal is determined for successive image areas. method. 6. Claim 5, wherein each image has at least 1000 such areas. The method described in section. 7. Claim 5, wherein each such area consists of at least 10 pixels. The method described in section. 8. 6. The system of claim 5, wherein each area is diamond-shaped. 9. A device for multiplexing the digital transmission signal with an analog video transmission signal. 2. The method according to claim 1, comprising: 10. 2. The method of claim 1, wherein the transmission includes a recording operation. 11. For transmitting and recording digital signals along with associated analog video signals. 11. The method according to claim 10, comprising a device for 12. Determine one or more motion vectors included in an image and calculate motion vectors for that image. comprising a device for transmitting the value of the motion vector in a digital transmission signal; , the method according to claim 1. 13. When a movement near one of the motion vectors is detected, the temporal or spatial Before deciding which filtering method should be used, the bandwidth compressor The method according to claim 12, taking into account movement in the amount of motion vectors. . 14. Two or more labeled motion vectors can be transmitted and controlled The signal is labeled motion relative to a continuous image area, where the image is stationary. moving in a specified one of the vectors, or in a manner that would be specified. The person according to claim 12, which is determined as indicating that the person is moving. formula. 15. The control signal generator includes a reconfiguration device similar to that in the receiver and a reconfiguration device similar to that in the receiver. Compare the operation of the reconstructor in each of its methods to determine the best method. 2. The method according to claim 1, comprising a comparison device for determining. 16. An application for the receiver to convert the interlaced transmitted video signal to a sequential display standard. up-converting device, and the up-converting device responds to the digitally transmitted signal. The method according to claim 1, wherein the method is adapted to: 17. video signal source, The bandwidth of the video signal generated by the source is divided into two or more different bandwidths in response to a control signal. a bandwidth compression device for selectively compressing the bandwidth by a method according to the invention; device for transmitting bandwidth compressed video signals in analog form; A bandwidth compressor is configured to transmit bandwidth to a bandwidth compressor in response to the image content of a video signal generated by a video signal source. a control signal that generates a control signal commanding the selection of the desired one of the width compression methods; a generator and a digital signal containing said control signal in conjunction with an analog video signal; equipment for transmitting signals; Video signal transmitter. 18. Transmission where bandwidth compression is performed selectively by two or more different methods an apparatus for receiving a bandwidth compressed video signal; Selectively display in two or more different ways depending on the image content from the received signal an apparatus for reconstructing a signal for, and a device for displaying the reconstructed signal; Receives digital signals transmitted in conjunction with video signals and generates control signals from them Extract and select the desired one of the reconstruction methods to the reconstruction device and send the signal for display. a device for commanding to reconfigure; Video signal receiver.