CN1041502A - The encoder and the decoder that are used for the transmission system of bandwidth reduction - Google Patents
The encoder and the decoder that are used for the transmission system of bandwidth reduction Download PDFInfo
- Publication number
- CN1041502A CN1041502A CN 89101692 CN89101692A CN1041502A CN 1041502 A CN1041502 A CN 1041502A CN 89101692 CN89101692 CN 89101692 CN 89101692 A CN89101692 A CN 89101692A CN 1041502 A CN1041502 A CN 1041502A
- Authority
- CN
- China
- Prior art keywords
- signal
- mentioned
- subsampling
- frequency
- sampling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Landscapes
- Television Systems (AREA)
Abstract
In 2: 1 compressed bandwidth reduction transmission systems of interlaced tv signal system bandwidth, encoder one side, carry out deviation subsampling (154 between the field, 156,162) and interframe deviation subsampling (160,164,168,170), from the interframe subsampling, can produce like this do not comprise anti-phase obscure the part the subsampling multiway transmission signal.During decoding, the composite transmission signal of subsampling is carried out interpolation handle (176,182,184,188),, simultaneously same signal is carried out between the field interpolation and handle (178,190), to obtain the active images part signal to obtain the still image part signal.Utilize frame-to-frame differences can detect amount of exercise, receiver circuit is simple and image quality improves.
Description
The present invention relates to a kind of encoder and decoder that is used for the transmission system of bandwidth reduction, be used to transmit broadband high sharpness colour TV picture intelligence, thereby this signal is suitable for narrow band transmission through readjusting.The invention particularly relates to the transmission system that also to finish the bandwidth reduction of motion detection by decoder rapidly satisfactorily.
The NHK(Japan Broadcasting Corporation) once proposed a kind of transmission system of bandwidth reduction, be used on a passage with 5: 3 width of cloth shape than broadcasting 1125 row HDTV(high definition TVs) image.This bandwidth reduction system is MUSE(multichannel Nyquist sampling coding), this system also is a kind of secondary sample system of compensating motion.
Below, will do one to muse system and briefly describe, at first, with sampling and the interpolation method in the narration muse system.
Compression bandwidth when being used for the high definition TV analogue transmission after the inferior Nyquist sampling method of paraphase combines with the inter-frame coding that is used for compensating motion, and people have developed the equipment that is used for 1125 row systems.
Table 1 has provided the most important characteristic of muse system, and Fig. 1 shows the systematic sampling pattern.Be sampled as every the multi-point scanning pattern, and be one-period and sequential loop with four.
For rest image zone (the image freeze part of field), can use the sampling of four field signals to reconstitute the HDTV image by instantaneous interpolation.The zone that shows the spatial frequency range of understanding the rest image district among Fig. 2 B.
For moving image,, utilize the sampled signal of a field can constitute final image by spatial interpolation.If the signal with two or more reconstitutes motion video, then Tu Xiang technical quality is because multirow is fuzzy will reduce.
By means of the spatial interpolation method, compressible transmission area, shown in Fig. 2 B, this motion parts that shows image fuzzy image will occur because of a unlapped background.But, since human eye to fuzzy in the picture motion part almost experience less than, so quality descends and is not serious.
Table 1
The characteristic of muse system
Multiple subsampling (the C of system of system balance motion
The multiplexing of signal is the TCI form)
Scan 1,125,/60 2: 1
The bandwidth 8.1MHZ(-6dB of transmission base band)
Secondary sample clock rate 16.2MHZ
The stationary part of horizontal bandwidth (Y) 20-22MHZ(image)
12.5MHZ*(the movable part of image)
(C) stationary part of 7.0MHZ(image)
3.1MHZ*(the movable part of image)
Positive digitalization is synchronous synchronously
Audio frequency and additional information be multiplex PCM in VBLK,
Adopt 4 phase DPSK(2048Kb/s)
* the value of prototype receiver: if adopt best two-dimensional digital filter, these values should be 16MHZ and 4MHZ.
When causing motion owing to pitching and panoplay camera lens, even more serious fuzzy with producing, for avoiding this influence of spatial interpolation method, introduced motion compensation.By encoder, can the vector of every representative shot motion be calculated, and vector signal is multiplexed to receiver in vertical blanking period.In decoder, the position of moving the pixel of previous field sampling according to motion vector with compensating motion, can be used for pitching and panoplay camera lens with instantaneous interpolation and that what do not have is fuzzy.Shown in Fig. 2 B, it is that the stationary part maximum can be transmitted half of field frequencies range that the maximum that is used for visual movable part can be transmitted field frequencies range, its reason is that former HDTV signal is interlacing scan in 2: 1, if the spatial interpolation method is used to the stationary part of image, then maximum can be transmitted a spatial frequency and will double and equal 1/2h, the space of h between this expression horizontal sweep.
Secondly, descriptive system is constituted.Fig. 3 A and 3B show the block diagram of understanding the MUSE transmitter and receiver.At first, the HDTV vision signal becomes the TCI signal by TCI encoder 2 codings.Fig. 4 shows clear TCI waveform with capable forsequential color signal.The sampling frequency of TCI signal is 64.8MHZ.Before this signal carried out subsampling with 16.2MHZ, they were respectively applied for static and moving areas prefilter 4 and 6() obtain input according to the situation of image area.These two filters 4 and 6 ideal characterisitics are shown in Fig. 2 A and 2B.
8 pairs of filters 4 of blender and 6 output mix.Its mixing ratio depends on the motion of image, and it will be carried out detection by pixel.Mix output and carry out subsampling by subsampling circuit 10.Some are combined by multiplexer 12 and subsampling signal such as the control signal of motion vector one class.The MUSE signal of combination is then by 14 modulation of FM modulator.
By the PSK(phase shift keying) modulator 22 modulates 4 phase DPSK(differential phase shift keyings with audio signal) signal.This signal after the modulation of MCSE signal frequency, utilizes field blanking interval and vision signal to carry out multiplexing by switch 15.Transmission and carry out multiplexing in the control signal blanking interval on the scene with baseband signal.
Shown in Fig. 3 B, in receiver, the signal of reception by FM demodulator 24 and PSK demodulator 26 rectification to obtain the Voice ﹠ Video signal of demodulation.Vision signal after the demodulation is separated video and the control output that separates with acquisition by demultiplexer 28.Video output adds to spatial interpolator 32 and instantaneous interpolater 34.At this, be static or motion according to image, and which interpolater decision adopt.That is the signal controlling blender 38 after, the moving region is detected and detected by detector 36.Add to the TCI decoder to obtain signal from the signal of 38 outputs corresponding to original video signal.
Blender 38 mixes 32 and 34 output.Blender 38 should be controlled by pixel, but in this case, the transfer rate of control signal is very high, thereby can not transmission signals.Therefore, must adopt the transmission signals of subsampling that motion is detected, and in the following muse system that NHK proposes, can accurately detect motion by receiver.
In muse system, carry out following motion detection, pixel is at motion parts or at the stationary part of image on earth, can detect by means of the signal difference with preceding frame.Owing to carried out subsampling,, still, but can obtain the difference between existing frame and the next frame, and it can be used for replace real difference so can't from the MVSE signal of transmission, obtain each frame difference.In some cases, the method does not provide real movable information, and as shown in Figure 5, label is that the motion parts of β can only can not detect from the signal of next frame from existing frame, the straightforward procedure that overcomes this difficulty is this difference of instantaneous expansion in next frame, as shown in Figure 5.
Concerning nearly all HDTV image, can adopt this method for testing motion, but some exceptions are also arranged, as the grid pattern with the specific speed panoplay, it has provided the partial images identical with preceding frame and has produced no movable information.
To this camera lens, adopted accurate frame-to-frame differences, it is poor between the former frame that obtains of existing frame and spatial interpolation method, certainly, image blurs.
In this way, the stationary part with image of high spatial frequency component may be judged as motion parts.The application of accurate frame difference is limited to the part that can not detect motion with second frame-to-frame differences.According to multiplex obstruction control signal, judge which kind of method static and motion (activity) image area are adopted.
Yet, with regard to the decoder of muse system, but in the detection of active images, exist a problem, promptly require by distinguishing that active images part and still image partly remove to handle signal.That is, in muse system, the subsampling cycle comprises two frames, so when detecting motion, can not use frame-to-frame differences (because do not exist in a subsampling in the cycle target with the detected motion difference).Its result must detect the difference between " inferior consecutive frame ", so motion detection is unsatisfied.In this manual, term " frame-to-frame differences " be meant for example signal level difference between first and second frames, and " difference between the inferior consecutive frame " is meant for example signal level difference between the first and the 3rd frame.
Why bad reason will be described in detail later motion detection.
Corresponding to the still image part, can carry out interpolation by adopting the signal in the frame, and this interpolation can not be used for the active images part.Thereby,, can realize the interpolation of active images part by adopting the signal in the frame.Because the interpolation method of these different modes needs to handle cutting apart between static and the moving region.
Therefore, to decoder, must with high accuracy movable image information be detected according to the picture intelligence of transmission, and in muse system, sampling frequency comprises two frame periods as mentioned above.Therefore, must be between two time consecutive frames the information of detected activity image, so motion detection comes down to incomplete.
Can see above-mentioned relation from following different viewpoints.This is second sampling frequency with 16MHZ(then with 32MHZ sampled (this is first sampling frequency) to suppose to have the signal of frequency spectrum shown in Fig. 6 A) carry out subsampling.Then shown in Fig. 6 B and 6C, transmit the high fdrequency component (8MHZ～24MHZ) be confused of base band.In this case, become false relation between low high fdrequency component is mutual certainly, so that avoid overlapped.Used among Fig. 6 C " at the interframe homophase " speech is meant this a kind of situation: promptly when high fdrequency component was obscured, (amplitude as 8～12MHZ) was synchronous to the corresponding signal of each frame subsequently.Similarly definition is used to " homophase between the field ".
But, because the result of interframe skew subsampling, i.e. the result of subsampling, the phase place of the amplitude of the high fdrequency component in each frame is anti-phase 180 ° subsequently.So can not from the waveform shown in Fig. 6 C, obtain " frame-to-frame differences ".As a result, must from the signal between two frames, ask for action message, and the amplitude of the high fdrequency component in this two frame has same phase.
In this manual, " interframe/in the ranks the depart from subsampling " speech shown in Fig. 6 C be meant utilization at every frame and every row at (for example point in the 4n field of Fig. 1 and (4n+2)) clock on the corresponding sampling point by subsampling that paraphase realized.
" between depart from sampling " speech be meant utilization to every clock by the sampling that paraphase realized, for example, these sampling points are corresponding in the 4n field with at the sampling point of (4n+2) with corresponding to the sampling point in (4n+1) and (4n+3) field.
According to as mentioned above, one of purpose of the present invention is that a kind of encoder and decoder that is used for the bandwidth reduction transmission system will be provided, and wherein, for decoder, can fully detect motion by means of frame-to-frame differences.
Another object of the present invention provides a kind of encoder and decoder that is used for the bandwidth reduction transmission system, and it can make receiver oversimplify and also improve image quality greatly.
Further purpose of the present invention provides and can correctly detect the movable decoder of little image.
A further object of the invention provides a kind of decoder with motion detector of simple structure.
Another object of the present invention provides a kind of decoder, and it can correctly detect the motion relevant with colour signal.
An object of the present invention is to provide a kind of decoder, it can be eliminated MUSE-II transmission signals is reduced the processing of noise and the movable inhomogeneities of image that produces.
Further purpose of the present invention provides a kind of decoder, and it can eliminate the movable inhomogeneities of image that the decoding of MUSE-II transmission signals is produced.
The present invention also has a purpose to be to provide a kind of decoder, and it can prevent the image diffusion of resetting, because have between the field of reducing number of taps or the intrafield interpolation method can not show Ideal Characteristics.
The present invention also aims to provide the decoder with interpolation filter between the field, this filter can reduce hardware size.
By forming the subsampling transmission signals, can obtain above-mentioned and other purposes of the present invention, described transmission signals does not comprise the part of obscuring in the low frequency range.
In a first aspect of the present invention, a kind of by departing from subsampling between realizing and interframe departs from the encoder that subsampling compresses the bands of television signals of 2: 1 interlaced scanning systems, comprise: first sampler, it is to have the timing sampling pulse corresponding to first sampling frequency of the upper limiting frequency that is essentially vision signal, the vision signal of TV signal is carried out departing from subsampling between the field: comprise a low pass filter, it receives the output signal of first sampler and departs from the signal of obscuring part that subsampling produces between the field by having, and the cut-off frequency of low pass filter is substantially equal to 1/2 of first sampling frequency; Comprise one second sampler, it receives the output signal of low pass filter, and with timing sampling pulse this output signal is carried out interframe and depart from subsampling with second sampling frequency, this second sampling frequency is lower than first sampling frequency but is higher than half of first sampling frequency, like this, do not comprise that the anti-phase multichannel subsampling transmission signals of obscuring part can depart from subsampling by interframe and obtain.
Here, carry out departing between the field first sampling frequency of subsampling and the ratio that carries out interframe and depart from second sampling frequency of subsampling in the ranks can be set at 3: 2.
In a second aspect of the present invention, the encoder of the bands of television signals that transmission is compressed by departing from subsampling and interframe to depart from subsampling between the field, comprise: prefilter between, it receives an incoming video signal and transmits the signal that has corresponding to the first transmission characteristic bandwidth; Comprise first sampler, be used for receiving the output signal of first sampler and regularly this output signal carried out departing from subsampling between the field with sampling pulse with first sampling frequency; Low pass filter, it receives the signal of first sampler output and departs from the signal component of obscuring part that subsampling produces between the field by having, and the cut-off frequency of this low pass filter is substantially equal to first sampling frequency half; Comprise first converter apparatus, the output signal that it receives low pass filter also converts the sampling frequency of this output signal to different sampling frequency; Comprise prefilter between a field, receive the vision signal of input and transmit the signal that has corresponding to the second transmission characteristic bandwidth; Comprise one second converter apparatus, the output signal of prefilter also converts the sampled signal of this output signal to different sampling frequency between received field; Comprise a detector means, it detects visual activity in response to incoming video signal; Comprise a mixer apparatus, be used for according to the output signal mixing of the output signal of activity detection apparatus with first and second transducers; And comprise second sampler, it receives the output signal of mixer apparatus and with the sampling pulse timing with second sampling frequency this output signal is carried out interframe and depart from secondary sample, described second sampling frequency is lower than first sampling frequency but is higher than 1/2 of first sampling frequency, so do not comprise that the anti-phase multichannel subsampling transmission signals of obscuring part can depart from subsampling from interframe and obtain.
At this, first sampling frequency that departs from subsampling between with interframe and the ratio that departs from second sampling frequency of subsampling in the ranks can be set to 3: 2.
In a third aspect of the present invention is decoder, and it comprises: receive the device of multichannel subsampling transmission signals, this transmission signals does not comprise that interframe is obscured part in the low frequency component; First interpolation device is used for that multichannel subsampling transmission signals is carried out interpolation and handles to obtain the signal of still image part; Second interpolation device is used for that multichannel subsampling transmission signals is carried out interpolation between the field and handles to obtain the signal of active images part; Motion detection apparatus detects to obtain the momental motion detection signal of presentation video in order to the frame-to-frame differences to multichannel subsampling transmission signals; Mixer apparatus, it is used for according to picture motion amount from the motion detection signal that motion detection apparatus obtains, and the signal of the active images that obtain to the still image signal that obtains from first interpolation device with from second interpolation device carries out linear hybrid.
Here, first interpolation device can comprise the frame interpolation device that multichannel subsampling transmission signals is carried out frame interpolation, and comprises multichannel subsampling transmission signals is carried out interpolation device between the field of interpolation between the field.
Mixing arrangement can comprise first mixing arrangement and second mixing arrangement.The size of the picture motion amount in the motion detection signal that first mixing arrangement is sent according to motion detection apparatus, the linear mixing of an output signal that the output signal that the frame interpolation device is sent and second interpolation device are sent; Interpolation device is realized interpolation between the field to a signal of sending here from first mixing arrangement between; Second mixing arrangement is according to the picture motion amount in the motion detection signal, with a signal and a signal linear hybrid of moving from interpolation device between the field of being sent here by first mixing arrangement.
Motion detection apparatus can comprise:
A frame-to-frame differences checkout gear is used for detecting the frame-to-frame differences in the multiple subsampling transmission signals;
Low pass filter, this device have two selectable cut-off frequencies, and this filter is used for taking out motion detection signal in the frame-to-frame differences signal of being sent here by the frame-to-frame differences checkout gear; One comprises folding (obscuring) part in two cut-off frequencies, and one does not comprise folding (obscuring) part;
A choice device is used to select one of them of two cut-off frequencies of low-pass filter;
Checkout gear is used for detecting the difference between adjacent two frames of multichannel subsampling transmission signals;
Output device, this device is according to coming self-test device, obtaining a compensating signal corresponding to inferior adjacent other difference signal of two frame-to-frame differences, this compensating signal is sent to choice device, makes choice device select suitable one according to this compensating signal in two cut-off frequencies.
Output device can contain an inhibiting apparatus, and this device is in response to the frame-to-frame differences signal that postpones a frame with respect to multichannel subsampling transmission signals, in order to the difference signal of the inferior adjacent interframe of forbidding being sent here by checkout gear.Can be controlled like this choice device: make its when inhibiting apparatus is forbidden difference signal not comprising that frequency of folded part elects the cut-off frequency of low pass filter as.
Interframe difference checkout gear can comprise anti-phase device, and the signal that this device is sent the frame interpolation device here every half clock signal period (clock signal frequency is an interframe subsampling frequency) carries out anti-phase.
Interframe difference checkout gear can also comprise the device of removing the folded part in the multichannel subsampling transmission signals.
The 4th part of the present invention, decoder comprises:
Receiving system is used for receiving the multichannel subsampling transmission signals that its low frequency component does not comprise the interframe folded part;
First interpolation device is used for that multichannel subsampling transmission signals is carried out interpolation and handles, to obtain the signal of still image part;
Second interpolation device is used for multichannel subsampling transmission signals is carried out interpolation processing between the field, to obtain the signal of active images part;
Motion detection device is used for multichannel subsampling transmission signals is carried out interpolation processing between the field, to obtain the signal of active images part;
The size of image motion amount in the activity detection signal that a mixing arrangement, this device are sent here according to activity detection apparatus, the active images signal that the still image signal that first interpolation device is sent here and second interpolation device are sent here carries out linear hybrid.
A signal processing apparatus, the signal of the output signal that this device is sent above-mentioned mixing arrangement here after as decoding delivered in the high fdrequency component of frequency of low frequency component that its frequency is higher than multichannel subsampling transmission signals, and the signal of multichannel subsampling transmission signals after as decoding sent in the low frequency component;
The signal processing apparatus here can comprise:
Separator is used for low frequency component and high fdrequency component are separated from multichannel subsampling transmission signals, and this separator is input to the high frequency signal of multiple subsampling transmission signals in first interpolation device and second interpolation device;
Low frequency component signal addition in the multichannel subsampling transmission signals that output signal that adding device, this device are sent mixing arrangement here and separator are sent here.
Signal processing apparatus can also comprise an alternative, and this device is replaced low frequency component signal in the mixing arrangement output signal with the low frequency component signal in the multichannel subsampling transmission signals.
Below in conjunction with the accompanying drawings in the description to optimum implementation of the present invention, above-mentioned and other purposes, effect, feature and advantage of the present invention will become more obvious.
Fig. 1. be a key diagram of having represented the sampling mode of common HDTV system.
Fig. 2 A and 2B have shown the spatial frequency zone that can transmit in the common muse system.
Fig. 3 A and 3B are the calcspars of transmitter and receiver in the common muse system.
Fig. 4. the form of transmission signals in the common muse system that drawn.
Fig. 5. be the sequential chart that is used for explaining the General Principle that movable part detects.
Fig. 6 A～6C is the transmitting pin spectrogram in the common muse system.
Fig. 7 A～7D is the spectrogram that is used for explaining basic principle of the present invention.
Fig. 8 A and 8B are respectively the block diagrams of an embodiment of the embodiment of an encoder of the present invention and a decoder.
Fig. 9 A and 9F are used for the encoder among key-drawing 8A and the 8B and the spectrogram of decoder mode of operation.
Figure 10 A is to use the block diagram of a specific embodiment of prefilter between a field in the present invention.
Figure 10 B is the schematic diagram that is used to explain preset filter between above-mentioned.
Figure 11 has shown the block diagram of the specific embodiment of preset filter between field that the present invention uses.
Figure 12 A and 12B are the block diagrams of two embodiment of linear mixer circuit of the present invention.
Figure 13 A and 13B are the block diagrams of two embodiment of moving detector circuit of the present invention.
Figure 14 is the block diagram of other embodiment of encoder and decoder among the present invention.
Figure 15 A～15E has shown the encoder that is used for explaining among Figure 14 and the sampling mode of decoder mode of operation.
Figure 16 is the block diagram of the other embodiment of decoder among the present invention and encoder.
Figure 17 A is the block diagram of the low pass filter 214 among Figure 16.
Figure 17 B is the key diagram that is used for the sampling frequency conversion of key-drawing 17A.
Figure 18 is the block diagram that receives a decoder in the receiver of MUSE-II transmission signals.
Figure 19 is the block diagram of a special embodiment of the moving detector circuit among Figure 18.
Figure 20 is a sequential chart, represented among Figure 19 or door output signal and produce relation between the transmission signals in the circuit of signal of change over switch S.
Figure 21 is the block diagram of the circuit of another signal that produces change over switch S3.
Figure 22 is a block diagram, has shown another embodiment of moving detector.
Figure 23 is the block diagram of another embodiment of moving detector.
Figure 24 is the block diagram of another embodiment of decoder that is used for receiving the receiver of MUSE-II transmission signals.
Figure 25 is the block diagram that receives another embodiment of the decoder in the receiver of MUSE-II transmission signals.
Figure 26 is the block diagram of an embodiment of the low pass filter among Figure 25.
Figure 27 is the block diagram of an embodiment of interpolation filter between the field.
Figure 28 is the block diagram of another embodiment of interpolation filter between the field.
Figure 29 is the circuit diagram of an embodiment of the variable coefficient filter among Figure 28.
Embodiments more of the present invention of describing below, it is the improvement that muse system above-mentioned is done, therefore the MUSE-II system that is called in the following description book, transmitter circuitry is referred to as " MUSE-II encoder " and receiver circuit is referred to as " MUSE-II decoder ".
The sampling mode of the MUSE-II encoder of above-mentioned introduction is identical in essence with sampling mode in the above-mentioned muse system of mentioning.
That is to say that in this MUSE-II system, offset sampling and interframe offset sampling are combined use between.Consequently, offset sampling has reduced oblique resolution between, makes amount of information be reduced to original half; The interframe offset sampling transmits an images with two frames, and the amount of information that therefore every frame comprised also is reduced to half.
Fig. 7 A～7D has shown an example according to coded sequence of the present invention.At first shown in Fig. 7 A, its bandwidth be limited in signal in the 24MH with the sampling rate of 24MHz by offset sampling between the field, the frequency spectrum that obtains is shown in Fig. 7 A.
Then, a low pass filter is removed the frequency component (shown in Fig. 7 B) that signal intermediate frequency rate after the above-mentioned sampling is higher than 12MHz.Then, the signal that obtains is carried out the interframe offset sampling with the sampling frequency of 16MHz.The result is that the frequency spectrum that obtains (does not comprise interframe folded part (its amplitude mutual opposite component in the homophase frame just) in base band component in 0～4MHz zone of 0～8MHz).Among Fig. 7 D, a low pass filter removes the frequency component that the signal intermediate frequency rate after taking a sample like this is higher than 8MHz.
Therefore, at decoder over there, after the signal of its frequency spectrum shown in 7D was the low pass filter of 4MH by cut-off frequency, decoder can not comprised the signal of folded part, can obtain " interframe difference " signal according to this signal, this frame-to-frame differences level signal can be used as activity detection.
According to the present invention, being sent out signal is the signal of its frequency spectrum shown in Fig. 7 D, this just the multichannel subsampling transmission signals of interframe folded part (this specification with the rear section in, this signal is called as MUSE-II transmission signals).Above-mentioned interframe folded part is that the signal that will transmit is carried out the interframe offset sampling in low-frequency range formed.
MUSE-II signal is carried out interpolation between frame interpolation and field, to reappear a signal corresponding to primary signal.According to the present invention, activity can detect according to the interframe difference, and therefore, live image part and still image part can make a distinction each other fully, can be improved so that reappear the image quality of image.
In when decoding, the signal shown in Fig. 7 B is corresponding to by frame interpolation but there is not the signal of 16MHz signal component.This means: when live image part was used as rest image mistakenly and is partly deciphered, no matter be that live image is partly gone back the rest image part, the calibration signal component of the 16MH of generation can be eliminated.Even the signal of frame interpolation acquisition just is the low pass filter of 16MHz by cut-off frequency, intrinsic information is unlikely to be lost.Even active images parts is used as still image mistakenly and is partly deciphered, can generation can not cause that the 16MHz of serious disturbance splits yet, therefore, image quality is improved greatly.
Next, we describe a more detailed embodiment of the present invention according to Fig. 8-15.
Fig. 8 A has shown the general structure of a MUSE-II encoder that sends MUSE-II signal.A/D converter of reference number 42 expressions among Fig. 8 A, transducer is used to obtain the input signal shown in Fig. 9 A.Reference number 44 is represented prefilter between a field, and its input characteristics is by shown in Fig. 9 B.Reference number 46 is represented a sample circuit, and this sample circuit carries out being offset subsampling between the field with the sampling frequency of 24.3MHz to the output signal of prefilter 44.Numeral 48 is represented a sampling frequency converter, and this converter becomes 32.4MHz with sampling frequency by 24.3MH and do not change the frequency domain (being frequency spectrum) of signal component.
Prefilter between field of numeral 50 expressions, its output characteristic is shown in 9D.Numeral 52 is represented a sample circuit, and this circuit carries out interframe skew subsampling to the output of prefilter 50 with the sampling frequency of 24.3MHz.Numeral 54 is sampling frequency converters, and its 26S Proteasome Structure and Function sampling frequency converter 48 is identical in essence.
Activity detector of reference number 56 expressions is in order to the movable part in the output that detects A/D converter 42.Linear mixer of reference number 58 expressions, it is according to the output signal of activity detector, being mixed corresponding to the output signal of still image converter 48 partly with corresponding to the output signal of active images converter 54 partly.
Sample circuit of reference number 60 expressions, this circuit carries out interframe skew subsampling to the output of blender 58 with the sampling frequency of 16.2MHz.62 is D/A converters, and it converts the output of subsampling circuit 60 to analog signal.Transmission filter of numeral 64 expressions, the signal component that this fluid strainer allows to be complementary with transmission line characteristics in the above-mentioned analog signal is passed through.
" prefilter 44 between " of foregoing description is a three-dimensional (trunnion axis, vertical axis and time shaft) filter, and what send into this fluid strainer is two field signals in succession.In addition, prefilter 50 is a two dimension (trunnion axis and vertical axis) prefilters between, sends into the just signal in of this filter.
Figure 10 A is the detailed diagram of an embodiment of prefilter 44 between field among Fig. 8 A.Among Figure 10 A, field memory of reference number 84 expressions, it postpones one (562 row) to the input signal from A/D converter 42.Numeral 86,88 and 90 all is a horizontal delay; 92 and 94 is adders; 96,98 and 100 be shift register; 102,104 and 106 be adder, they multiply by the results added that obtains behind weighted factor, β, the γ to numerous outputs of shift register respectively.108 also is an adder.The weighted factor that mentions the above value that they obtain when to be two-dimentional impulse response with the sampling frequency of 48.6MHz (vertical direction 1125 row) sampled itself, weighted factor, β, γ sum are 1.
Figure 11 is the detailed diagram of an embodiment of the interior prefilter 50 in field among Fig. 8 A.Here, reference number 110,112,114 and 116 represent that all retardation is the 1H delay line of delegation, they will postpone delegation successively from the input signal of A/D converter 42; 118 and 120 represent two adders, and they are respectively the mutual addition of output of the output of input signal among the figure and delay line 116, delay line 114; Numeral 122, three transversal filters of 124 and 126 expressions, these filters pass through the output of an adder 118,120 and a delay line, adder of reference number 128 expressions, this adder is added up transversal filter 122,124 mutually with 126 output signal.
Figure 12 A is a block diagram, and this block diagram shows the basic principle of the linear mixer 58 shown in Fig. 8 A.This linear mixer 58 comes mixed video signal A and B with mixing ratio α (0＜α＜1), wherein:
A: be the vision signal that its activity diagram pixel has been carried out processing;
B: be the vision signal that its still image element has been carried out processing;
α: the signal relevant with activity is under the situation of still image pixel at pixel, α=0, and when pixel is the active images pixel, α=1, this is 4 bit signals (0≤α≤1) of representing the various phenomenons between the static active images.
Circuit among Figure 12 A needs two multipliers 130 and 132, so its structure not only complexity but also costliness.Preferably adopt the circuit shown in Figure 12 B in the reality.
Among Figure 12 B, the output signal that subtracter 136 produces is (A-B), and multiplier 128 multiplies each other this signal and α; The output signal α of multiplier 138 (A-B) again with the B signal plus, the signal that obtains is B+ α (A-B)=α A+(1-α) B.
Figure 13 A and 13B are the The general frame of two embodiment of the activity detector 56 among Fig. 8 A.Among Figure 13 A and the 13B, frame memory of reference number 142 expressions, it is stored in the frame video signal from A/D converter 42 wherein; 144 is subtracters, is used for obtaining interframe difference (i.e. the level difference of two interframe) signal, and 146 is low pass filter/absolute value circuits, is used for receiving above-mentioned frame-to-frame differences level signal; 148A, 148B and 148C are the read-only memorys (ROM) of realizing nonlinear transformation; 150 is high pass filters, is used to detect the edge of incoming video signal.
Explain the operating state of the sort of MUSE-II encoder described above below in conjunction with 8A:
(1) A/D converter 42 is taken a sample to input signal with the sampling frequency of 48.6MH, thereby obtains to have an output signal of bandwidth shown in Fig. 9 A.Horizontal component is demarcated along transverse axis, and signal level is demarcated along the longitudinal axis.
(2) use shown in Fig. 9 B between 44 pairs of still images parts of prefilter handled, thereby eliminate image high fdrequency component obliquely.Filter 44 has first transmission characteristic shown in Fig. 9 B, this characteristic will by between the ceiling restriction of frequency of horizontal component of prefilter on the sampling frequency of 24.3MHz.
(3) be offset subsampling between the sampling frequency achieve frame of sample circuit 46 with 24.3MHz.The signal that consequently makes frequency be higher than 12.15MHz folds for folding center with 12.15MH, thereby obtains the signal of its bandwidth shown in Fig. 9 C.
(4) sampling frequency converter 48 is transformed to 32.4MHz with sampling frequency by 24.8MHz, and under these circumstances, the bandwidth of signal remains unchanged.That is to say, have only sampling frequency to change, and the frequency bandwidth of signal remains unchanged.
(5) for active images parts is handled, have second transmission characteristic shown in Fig. 9 D shown in the field in prefilter 5O will be by prefilter between the field the bandwidth of horizontal component be restricted to 12MH.
(6) after this, the output of 52 pairs of prefilters 50 of subsampling circuit is taken a sample with the sampling frequency of 24.3MHz.Because the bandwidth of the output signal of prefilter is restricted to 12MHz in (5), therefore will can not fold, thereby signal bandwidth remain unchanged (shown in Fig. 9 D).
(7) sampling frequency converter 54 is transformed into 32.4MHz with sampling frequency by 24.8MHz.This moment signal bandwidth remain unchanged in essence (shown in Fig. 9 D).
(6) and the variation that can not produce signal bandwidth in essence of the processing in (7), so sampling frequency can be directly converted to 32.4MHz by 48.6MHz in (5).
(8) at first activity detector 56 obtains frame difference signal (shown in Figure 13 A and 13B) between adjacent two frames, obtain the absolute value signal of this frame difference signal then by circuit 146, this absolute value signal is converted to a nonlinear motion detection signal by ROM148A or ROM148B and 148C, thus the output movement amount.
(9) blender 58 carries out linear hybrid to the active images that obtain in the still image that obtains in (4) and (7) according to the amount of exercise that obtains above.
(10) next, carry out interframe skew subsampling with the sampling frequency of 16.2MHz.Secondly, carry out interframe skew subsampling with the sampling frequency of 16.2MHz.As a result, still image (Fig. 9 C) and active images (9D) become shown in Fig. 9 E and 9F, turn back at the 8.1MHz place respectively or bend.Bandwidth shown in Fig. 9 D reaches 12.15MHz, therefore, folded part can not occur when frequency is lower than 4MHz.
(11) last, digital to analog converter 62 will convert analog signal to from the digital signal of subsampling circuit 60.In this case, signal just is transferred to transmission line by transmission filter 64, and filter 64 can obtain the cosine roll off characteristic when frequency is higher than 8.1MHz.
Shown in Fig. 8 B, in receiver, the said process of transmitter has been fallen to come.
(12) at first, A/D converter 66 is taken a sample again.In this case, Fig. 9 E is depicted as the signal bandwidth of still image, and Fig. 9 F is depicted as the signal bandwidth of active images.
(13) insert filter 68 by interframe and in such a way still image is handled, cause the pixel in the previous frame signal to replace not sampled pixel.Like this, just reappear the frequency spectrum shown in Fig. 9 C by the folded part shown in Fig. 9 E.In this manual, interframe is inserted and to be meant the process that obtains to insert signal by sampled value, and used sampled value is to obtain from subsequently frame.
(14) sampling frequency converter 70 is transformed into 24.3MHz with sampling frequency from 32.4MHz, but signal bandwidth remains unchanged shown in Fig. 9 C.
(15) and, insert filter 72 between utilizing, so that go to reappear signal bandwidth shown in Fig. 9 B by folded part shown in Figure 9.
(16) insert filter 74 between use and handle active images, like this, just can reappear folded part shown in Fig. 9 D by the frequency spectrum shown in Fig. 9 F.
(17) after this, sampling frequency converter 76 just is transformed into 48.6MHz with sampling frequency from 32.4MHz.In this case, signal bandwidth remains unchanged shown in Fig. 9 D.
(18) activity detector 78 to 4MHz, to obtain frame difference signal, carries out Nonlinear Processing to this difference signal, to obtain an activity with the input signal bandwidth constraints again.
(19), still image and active images are mixed mutually linearly according to the activity that is obtained.
(20) last, produce an analog signal from D-A converter.In this case, still image has the signal bandwidth shown in Fig. 9 B, and active images have the signal bandwidth shown in Fig. 9 D.When a pixel a bit moves, between the signal bandwidth shown in Fig. 9 B and 9D, can obtain a signal band.
Figure 14 shows that a kind of actual modified version of encoder shown in Fig. 8 A and Fig. 8 B and decoder.In Figure 14, the frequency that contiguous arrow marks is a sampling frequency.
In MUSE-II encoder, the analog-digital converter that reference number 152 expressions are sampled to input signal with the frequency of 48.6MHz.Prefilter between reference number 154 expression is added on this prefilter from the output signal of A/D converter 152.The output signal of skew sub-sampling circuit 156 sampling prefilters 154 between.From the output signal of sample circuit 156, be 12MHz by low pass filter 158(cut-off frequency), be added on the sampling frequency converter 160.
The output signal of A/D converter 152 is also added on the prefilter 162 of showing up.The output signal of prefilter 162 is added on the sampling frequency converter 164.The output signal of sampling frequency converter 164 is added on the activity detector 166, obtaining a frame difference signal, and this difference signal is added on the linear hybrid circuit 168.
The output signal that hybrid circuit 168 receives from sampling frequency converter 160 and 164 according to the amount of exercise of representing from the frame-to-frame differences signal form, is mixed these output signals linearly.The output signal of linear mixer 168 is added on the interframe offset sampling circuit 170.
Sample circuit 170 is with offset sampling between the sampling frequency achieve frame of 16.2MHz.Utilize number-transducer 172 output signal of should taking a sample to be converted to analog output signal.Then, by transmission line 173 these digital output signals of transmission.
Here, 152,154,156,160,162,166,168,170 and 174 each several parts are similar to 44,44,46,48,50,54,56,58,60 and 62 each several parts among Fig. 8 A.
Modified model shown in Figure 14 is different from the MUSE-II encoder shown in Fig. 8 A, in this modified model, economize between field and be offset subsampling except active images, because defined bandwidth, so do not needed between the field signal bias subsampling by prefilter between the field 162.That is to say that the composition of frequency does not a bit change.
Because be added to the output that the input signal of activity detector 166 comes prefilter 162 between self-fields, so improved circuit in fact as shown in figure 14 is more satisfactory.
MUSE-II decoder among Figure 14 has analog-digital converter 174,174 and receives signal by transmission line 173 transmission, thus, and with the analog signal of transmission frequency sample, to form digital signal with 16.2MHz.Insert on the filter 178 inserting between this digital signal extra show between filter 176 and field.The output signal of these two filters 176 and 168 is added to respectively on the low pass filter 182 and sampling frequency converter 190 of 12MHz.The output signal of filter 182 is added on the sampling frequency converter 184.These two sampling frequency converters 184 and 190 are transformed into 48.6MHz with sampling frequency from 32.4MHz under the situation that does not change signal bandwidth.The output signal of sampling frequency converter 184 add show up between the skew adopt circuit 186 again, like this, just with adopting again that the sampling frequency of 24.3MHz carries out being offset between the field.The output signal of adopting again added to show up interleave on the filter 188, the output signal of analog-digital converter 174 also is added on the activity detector 192, to obtain the frame difference signal of expression activity.The output signal of filter 188 and transducer 190 is added on the blender 194, and like this, these output signals are just mixed according to activity.The output signal of blender 194 is added on the D-A converter 196, reappears output signal to obtain one.
MUSE-II decoder circuit shown in Figure 14 is similar substantially to the circuit of the embodiment shown in Fig. 8 B.
Figure 15 A～15E is used for the sampling structure schematic diagram of the MUSE-II encoder working method shown in key diagram 15A～15E, mark O place quilt, and sampling is not sampled (but giving interpolation in decoder) and mark the X place.
Shown in Figure 15 A, A/D converter 152 is taken a sample with the frequency of 48.2MHz, and sampled signal does not comprise folded part certainly.
Next, under the still image situation, by prefilter between the field 154, with the sampling frequency of 24.3MHz (be half (1/2) of the sampling frequency of previous 48.6MHz] carry out between be offset subsampling, like this, must obtain the sampling point shown in Figure 15 B.
After this, low pass filter 158 is by the frequency that is higher than 12MHz, just thus the point shown in the X among Figure 15 B is inserted, thus sampled data shown in Figure 15 C obtained.This that is to say, low pass filter 158 is got back to the sampling frequency that the output signal of 48.6MHz is taken a sample again, so, sampling frequency converter 160 is used for converting this output signal to shown in Figure 15 A sampling mode.In this case, only carried out the conversion of sampling frequency, and the composition of image or vision signal itself remains unchanged still.
Under the active images situation, obtain to have the data of 32.4MHz sampling frequency with basic and the similar method of said method.
In the MUSE-II decoder shown in the bottom of Figure 14, used the reverse order of various insertion filters 176,178,188 with the signal processing of carrying out in the encoder that coexists, carry out signal processing.This signal processing had been described with reference to Fig. 8 A and 8B and 9A～9F, like this, in this specification, just this was not described in more detail.
Figure 16 shows that according to another MUSE-II encoder embodiment of the present invention.In this embodiment, MUSE-II encoder is so design and constitutes, even it is with low-frequency clock work, so that the hardware size compactness.That is to say, handling under the situation of luminance signal, carry out the activity vector compensation corresponding to the one-period of 64.8MHz, then, by prefilter between a three dimensional field, is sampling frequency with 24.3MHz, carries out offset sampling between the field of static or motion video.Under the active images situation, signal obtains from two-dimentional prefilter, and the interframe offset sampling that is obtained is to carry out with the sampling frequency of 16.2MHz.
Still with reference to Figure 16, reference number 202 is represented low pass filters (being lower than 20MHz), 204 and 222 expression A/D converters; 206 expression clock cycle with frequency 64.8MHz were carried out the vector compensation circuit of vector compensation; Prefilter between three dimensional field of 203 expressions; 210 expressions, the second subsampling circuit; 212 and 218 expression low pass filters (being lower than 12.15MHz); 214 and 220 expressions are used for the low pass filter (their detailed circuit is shown in Figure 17 A-17C) of 48MHz to 32MHz conversion; Motion vector detection of 216 expressions; Capable preface transducer of 224 expressions; The 226 express time axial compressions circuit (X1/4) that contracts; TCI encoder of 228 expressions; Two-dimentional prefilter of 232 expressions, frame memory of 234 expressions; Marginal detector of 236 expressions; Motion detector of 238 expressions; Blender of 240 expressions; One second subsampling circuit of 242 expressions; Synchronous and the control signal circuit of a combination of 244 expressions; Control-signals generator of 246 expressions; And D/A converter of 248 expressions.
Low pass filter 214 or 220 a enforcement are shown among Figure 17 A and the 17B.Here, reference number 260,262 and 264 expressions are by the clock frequency of 48.6MHz d type flip flop that drive, series connection.The output signal of trigger 260,262 and 264 is added to respectively on the d type flip flop 266,268 and 270 that is driven by the clock frequency of 16.2MHz.The output signal of trigger 266,268 and 270 is added to respectively on d type flip flop 272,274 and 276.
Output signal A, B, C, D, E and the F of trigger 276,270,274,268,272 and 266 are added to respectively on adder 278,280,282 and 284, to obtain output signal (A+D), (B+C), (C+E) and (B+F).Be added to respectively on factor multiplier 286,288,290 and 292 the output signal (A+D) of adder 278,280,282 and 284, (B+C), (C+E) with (B+F), and the output signal D of trigger 268 is added on the factor multiplier 294.Multiplier 286,288,290,292 and 294 has multiplication factor α, α, α, α and α respectively.These factors α-α obtains from impulse response frequency sample, the 16.2MHz low pass filter with 97.2MHz.
Below, will make brief description to the working method of embodiment shown in Figure 16.
Based on the basic principle of above-mentioned Muse-II system, in this case, by behind the low pass filter 202, with the frequency of 48.6MHz brightness signal Y is taken a sample by A/D converter 204 in brightness signal Y.Here, the bandwidth of input signal is restricted to by low pass filter 202 and is lower than 20MHz.
After this, under the control of the motion vector detection that motion vector detection carries out, finish vector compensation corresponding to the one-period of 64.8MHz by the vector compensation circuit 206 that is operated in 64.8MHz.Circuit 206 can consist of a filter, and this filter then according to the speed of 48.6MHz, is weighted in the position of 64.8MHz, thereby can avoid carrying out A/D conversion.
In order to utilize subsampling circuit 210 to carry out being offset subsampling between the field of still image signal, will be added to from the output signal of vector shift circuit 206 between three dimensional field on the prefilter 208.After subsampling circuit 210 carried out initial sampling, the bandwidth of signal was restricted to by low pass filter and is lower than 12.15MHz.
Accept frequency inverted from the output signal of low pass filter 212 by 48.6MHz to 32.4MHz.Usually,, at first it is converted to 97.2MHz, and makes it, just can obtain 32.4MHz thus by prefilter for 48.6MHz is converted to 32.4MHz.
The motion video signal that obtains from the compensation motion vector circuit of carrying out corresponding to the vector compensation of the one-period of 64.8MHz 206, add to the low pass filter 220 that carries out 48MHz to 32MHz conversion by low pass filter 218, just can finish the conversion that becomes 32.4MHz speed thus.
After this, in order to carry out the interframe encode skew subsampling (second subsampling) of signal in secondary subsampling circuit 242, signal at first will be added on the two-dimentional prefilter 232 by before the TCI encoder 228.The output signal of prefilter 232 is added on the blender 240.
By being used for the low pass filter 214 of 48MHz to 32MHz conversion, will move or still image signal is transformed into 32.4MHz, the output that will change is added on the blender 240 by TCI encoder 228 then.
The output signal A of TCI encoder 228 is added to frame memory 234 and motion detection 238.The motion vector detection output signal of frame memory 234 origin autokinesis vector detection devices 216 is controlled, and the output signal of frame memory 234 is added on the motion detector 238.Also the output signal of prefilter 232 is added on the marginal detector 236, and the output signal of detector 236 also is added on the motion detector 238, like this, motion detector 238 just produces a motion detection signal of wherein considering the motion parts edge.The mixing ratio of 240 two input signals of this motion detection signal control blender.
Colour signal C is added to time shaft compressor circuit 226(X1/4 by A/D converter 222 and row preface transducer 224), and carry out the time shaft demultiplexing with luminance signal with TCI encoder 228.Subsequently, use the mode identical to handle this colour signal with handling luminance signal.
Synchronous and control signal circuit 244 by combination is in the same place the motion vector detection signal controlling of control signal circuit 244 origin self-detectors 216 to signal that obtains from second sample circuit 242 and the signal combination that comes from control-signals generator 246.And be an analog signal by the conversion of signals that D/A converter 248 will come from secondary sample circuit 242.
Above-described control signal comprises the signal shown in those tables 2, and is detected by decoder.
Table 2
Control signal
The figure place content
(lowest order) 1 is corresponding to one of 64.8MHz
An amount in cycle
2 corresponding to of 32.4MHz
An amount in cycle
3 horizontal anomalous movement vectors are corresponding to one of 16.2MHz
An amount in cycle
4 corresponding to of 8.1MHz
An amount in cycle
5 directions (sine)
6 ± 1 row
7 vertical movable vectors ± 2 row
8 directions (sine)
9 brightness sampling phases
10 color sampling phase places
11 noises reduce control
12 11, the lowest order side
13 transmission coupling operating characteristic positions
The control of 14 activity detection
(H: muting sensitivity)
15 blank
16 000: static fully
17 movable conditioned signals 001: accurate static
010: normal
(highest order) 18 011-111: force the space to be inserted
An embodiment at the receiver side decoder that receives above-mentioned MUSE-II transmission signals shown in Figure 180.
In Figure 18, reference number 400 expression control signal/synchronizing indicators, it produces synchronizing signal according to the MUSE-II transfusion signal (the interframe sampling frequency with 16MHz) that receives at signal input part 402.Control the synchronous of whole decoder with the synchronizing signal that is produced.
S represents a subsampling shift switching, and it will be inserted in the transmission signals that is added on the input 402 from a signal of frame memory 404, with produce one wherein interframe inserted the signal of 32MHz sampling frequency.Frame memory 404 will postpone a frame from the input signal of end 402.This frame is that point " A " in Figure 18 obtains to the signal of the sampling frequency of inserting.When switch S is not conventional simple switch, be when being suitable for obtaining frequency and being the switch of 16MHz signal with suitable mixed proportion, just can guarantee reducing of noise.
The signal of switch S (it is the frame interpolation signal of 32MHz that sampling frequency is wherein arranged) is added to one of them in three inputs of motion detector 408, and also adds by switch S on the input of the blender 412 that interleaves circuit 410 of showing up.Switch S is synchronoused working with switch S.Inserting circuit 410 between can be made of a two-dimensional low-pass filter.
Being added on other input of activity detector 408 from the transmission signals of 402 inputs with from the output signal C of frame memory 404, motion detector 408 responds the signal that these three input signals produce representatives activity detection in the image of transmission signals transmission successively.
Switch S only produces the signal of representative from the signal of the switch S sampling point in existing, and the signal that is produced added to show up interleaves on the circuit 410.The output of circuit 410 is added to again on the blender 412.
Insert between that signal that circuit 414 will result from one of the delay of field memory 416 inserts or interpolation to from blender 412 signals, like this, just obtain to insert signal between field that a sampling frequency is 48MHz with circuit 414.Be added to one of two inputs of sampling frequency converter/blender 420 inserting signal between thus obtained field.
Is the sampling frequency from blender 412 other input that the signal of 32MHz is added to sampling frequency converter/blender 420, like this, this signal just is converted into the sampling frequency that sampling frequency is 48MHz, and, it is mixed with the output signal of inserting circuit 414 between the field according to the visual activity that is obtained as the activity detection signal.What the output signal of sampling frequency converter/blender 420 was added to that time compression polymerization decoder 422(refers to is exactly hereinafter " TCI decoder ").422 convert this input signal to produce from output 424 required TV signal.
Figure 19 shows that an embodiment of activity detector shown in Figure 180.
In Figure 19, interframe movement detector of reference number 426 expression detects moving between first frame and the 3rd frame with it, the signal A shown in Figure 18 and C is added to its input respectively.Comparison circuit of reference number 428 expressions is used for limiting the output signal of self-detector 426 with an appropriate threshold Vth, so that obtain the signal of a bit.
Frame memory of reference number 430 expressions is used for storing the signal D from threshold circuit 428 during a frame; The output of threshold circuit 428 and frame memory 430 all is added to or the door 434.According to from or door 434 output signal, come driving switch S.
Frame memory of reference number 436 expression is used in during the frame storage from the input signal " B " of input 402.At the input signal shown in Figure 18, with 16MHz the transmission signal of sampling frequency just, it not only is added to frame memory 436, also be added to one of them input of subtracter 438.The output of frame memory 436 is added to another input of subtracter 438.
As a result, subtracter 438 produces a frame difference signal, is added to one of them input of switch S by first low pass filter 440, and filter 440 has a wide frequency characteristic, such as, signal level decline 6dB when 8MHz.Second low pass filter 442 has a narrow rate band, and it is connected with first low pass filter 440, and its output is connected to another input of switch S.First and second low pass filters 440 and 442 produce a filtrate characteristic that does not respond together more than high frequency 4MHz.
The output signal that obtains from switch S is added to absolute circuit 444, forms the output signal of representing the input signal absolute value successively at this.
Among this motion detector shown in Figure 19, when the difference between two frames indicated motion detection, the utilization with low pass filter 440 of wide bandwidth made in the frequency band that is mixed into frame difference signal of the component higher than 4MHz.As a result, make and to produce the motion detection signal that can detect the slight movement in the image fully.
Go back with reference to Figure 18, signal C includes a signal component that postpones two frames with respect to signal B.Be added to the transmission signal of frame memory 404 by switch S from input 402, by the circuit cycles formed by frame memory 404 and switch S twice.Therefore, signal C includes the signal that postpones a frame for signal B, and postpones the signal of two frames for B, and they all are to be the point sequence of speed with 32.4MHz.Signal A comprises signal B and the relative signal that postpones a frame in signal B that with 32.4MHz is the capable form of point sequence of speed.
Representative from the signal between inferior consecutive frame of threshold circuit 428, and is postponed the signal of exporting from frame memory of a frame for the difference signal between this representative time consecutive frame backward, all be added to or 434, so that avoid misjudgment.This mistake promptly is for coming starting switch S and the image of a motion is thought by mistake be a static image according to the difference between time consecutive frame.
Figure 20 represent to transmit signal with or door 434 output signals between relation, or the circuit at door 434 places is to be used to produce the circuit shown in Figure 19 of signal of starting switch S.As shown in Figure 20, transmit the position of signal indication along with the disappearance picture motion of time.Transmit signal according to this, the output signal that difference between time consecutive frame and the difference between the inferior consecutive frame can be used as threshold circuit shown in Figure 19 428 in the frame is obtained in front, between these difference signals or a door output signal can be obtained.This or a door output signal comprise one for interframe difference unnecessary portions.Therefore, when the very fast motion of a big object, some side effect just is left in the output signal of decoder shown in Figure 180.
From the viewpoint of image quality,, still a circuit embodiment that is used to eliminate this side effect is shown among Figure 21 although this side effect does not produce any serious problem.
Refer now to Figure 21, threshold circuit of reference number 446 expressions; 448 and 451 is the frame storage; 452 is INHIBITORY-gate; 453 be one or, its receives the output signal of INHIBITORY-gate 452 and from the signal of the frame of delay of frame memory 448, the switch S shown in deactivation Figure 19.The non-inverting input that is added to INHIBITORY-gate 452 from the signal (seeing Figure 19) of threshold circuit 428 outputs as the frame-to-frame differences level signal between time consecutive frame.
The signal shown in Figure 19 from subtracter 438 outputs, be added to an input of threshold circuit 446 as the frame-to-frame differences level signal, this threshold circuit compares input signal successively with an appropriate threshold, and the output signal of threshold circuit 446 is added to frame memory 448.The signal that has postponed a frame by frame memory 448 is added to INHIBITORY-gate 452.INHIBITORY-gate 452 also receives from the signal D of threshold circuit 428 output, removes to forbid the signal from 428 outputs of threshold frame memory.Signal from INHIBITORY-gate 452 outputs is added to frame memory 451.The output signal of INHIBITORY-gate 452 and frame memory 451 is added to or door 453.
According to signals that start in the above described manner or door 453 outputs, switch S is connected to the output of first low pass filter 440.
The purpose that difference detects between the inferior consecutive frame is the motion that helps to detect the image in the signal that is not detected by frame-to-frame differences.So when detecting the motion that transmits image in the signal by frame difference, the detection of difference is unwanted between the inferior consecutive frame.When a big object of which movement is very fast, must consider above-mentioned side effect.In this case, necessarily detected frame-to-frame differences.
In this motion detector, low pass filter has drawn from MUSE-II signal has the frequency component of frequency range from 0 to 4MHz.When from the signal of acquisition like this, accurately asking for frame difference, a frame memory must be set.
Yet,, do not make and obtained frame difference with this frame memory according to the circuit shown in Figure 22.Refer now to Figure 22, XOR gate of reference number 454 expressions is used to replace frame memory 436 shown in Figure 19 and subtracter 438.XOR gate 454 receives signal A shown in Figure 18 and the clock signal of 16MHz.
At the signal A shown in Figure 18, in the data of the data of this frame and former frame staggered in each pulse with frequency 32.4MHz.Therefore, when this signal is added to XOR gate 454, on each clock of 16MHz by anti-phase, thereby and obtained the to have broadband frame difference of (comprising the part of obscuring).Signal from XOR gate 454 outputs is added to first low pass filter 440.First and second low pass filters 440 and 442, switch S and absolute value circuit 444 all start with reference to figure 19 described modes with top.In this mode, can obtain frame difference.
When the sample mode considered in MRSE-II signal, can obtain motion detection signal by the circuit shown in Figure 23.In Figure 23, reference number 460 and horizontal scanning period of 462 expression 1H() delay circuit; 464 is an adder; 466 is one 1/2 multipliers; 468 for having the XOR gate of an inverting input; 470 is XOR gate; 472 is adders; 474 is one 1/2 multipliers; 476 is the low pass filters that cut-off frequency is 4MHz.
By the circuit shown in Figure 23, obtain one as frame difference comprise from signal A shown in Figure 180 obscure the part broadband signal, this is possible.
More particularly, the signal A shown in Figure 18 in vertical direction by anti-phase, and by adopt the circuit shown in Figure 23 come picked up signal A with this inversion signal and, so just can eliminate the double sampling influence.Its reason is, by anti-phase, therefore, when when vertical direction is looked, this frame and preceding frame are interleaved on each 1H the phase place of the sampled signal in the signal shown in Figure 180 on each 1H of vertical direction.
In this MUSE-II transfer system,, can not detect frame difference as colour signal at least now.Therefore,, can utilize the difference between time consecutive frame, but as known in prior art, this is impossible to detect a motion fully with regard to colour signal.
Therefore, preferably according to the frame difference that obtains together with the brightness letter, or, control colour signal according to the frame difference signal of such acquisition with inferior adjacent all differences between the frame that colour signal obtains together.This method is best actually, because in general colour signal and luminance signal have a high degree of correlation.
Figure 24 represents another embodiment of the present invention.According to this embodiment, can overcome by MUSEII and transmit the following problems that the decoding of signal causes.
That is to say, in MUSE-II decoder, require to have a frame memory.Therefore, realized that by this frame memory the noise of input signal reduces, thereby the required carrier power of conveyer line and the ratio (being referred to as " C/N " later on) of noise power can be reduced.Yet in general the noise of input signal reduces the quality that has reduced motion video.Therefore, in MUSE-II decoder, the influence that is subjected to the input signal motion parts is handled in muting minimizing.So, determine that this input signal represents a motion video or a still image, owing to noise thickens.Therefore, even when the flat image area of input imagery during in motion, also can the image error of this motion be judged as static image.As a result, the image area that is judged to be still image has been dropped on after its actual motion.Therefore, when gamma camera was moved, the motion of whole reception image became inconsistent.
Decode procedure in muse system also is so, and instantaneous insertion in the decoding that transmits signal, partly adopted at still image because of being in its source, and adopts insertion between the field in the image area of motion.As a result, the false judgment owing to input signal makes the flat image area of input signal drop on after its actual motion.
In addition, when MUSEII was transmitted signal interpretation, for whole image is obtained good quality, the detection sensitivity of input signal motion parts was low for well.As a result, make this false judgment be tending towards further increase.
On the other hand, under the situation that FM transmits or USB-AM transmits, noise all almost is contained in the high-frequency region of transmission signals, makes and need not reduce its noise level in low-frequency band.
What MUSE-II transmission signals was not included in low frequency region (than the low frequency district of MHz) interframe obscures (folding) part.We can utilize this situation to solve above-mentioned problem, and this problem is exactly:
(1) reduces the non-unification of the picture motion of the transmission signals that device handles by noise.
(2) non-uniformity of the picture motion that causes by the decoding of MUSE-II signal.
Except above-mentioned problem, still there is other problem in practice.That is to say, be designed for insert between realizing and the field of MUSE-II transmission signals between insert two dimensional filter the time, if the number of tap is few, it is impossible obtaining desirable two dimensional filter characteristic.As a result, the image of reception has been thickened.
The present invention can address these problems.
Figure 24 shows a basic circuit that can overcome above-mentioned these problems.In Figure 24, low pass filter of reference number 480 expressions; The 482nd, with reference to the described the sort of decoder of Figure 18, be provided with a motion detector in the outside of decoder 482; 484 is a sampling frequency conversion device; 48 6 is delay lines; 488 is a subtracter; 490 is a D flip-flop.The clock of 32MHz frequency signal is added to the input end of clock of trigger 490.Adder of reference number 492 expressions; S is a sub-sampling alteration switch.
For hybrid motion and static image, be created in motion detection signal used in the decoder 482 by sub-sampling alteration switch S, frame memory 404 and motion detector 408.
MUSE-II transmission signals is added to low pass filter 480, on subtracter 488 and the switch S.From MUSE-II transmission signals, obtained to contain the signal of the frequency component lower than 4MHz by this low pass filter 480.According to MUSE-II transmission signals that is added to "+" input and the output signal that is added to the low pass filter 480 of "-" input, subtracter 488 generates one with the high fdrequency component that is lower than the low frequency component complementation of 4MHz.The high fdrequency component that obtains like this is added to decoder 482.
In the signal of this reproduction TV, the low pre-component that is lower than 4MHz does not have by decoder 482, so make because the picture motion non-uniformity that decoding or noise minimizing cause can be eliminated, has therefore promoted visual stability.
The high fdrequency component of MUSE-II transmission signals (than the high part of 4MHz) is compared with the technology that has earlier and can be accessed stronger noise and reduce to handle, so can reduce desired C/N in transmission path.
In addition, in vertical direction, the low frequency component of the picture intelligence of reproduction (lower than 4MHz) by this filter, can not promoted vertical resolution so significantly.Therefore, even because the incomplete insertion of the high fdrequency component (higher than 4MHz) of the picture intelligence that produces again in decoder when vertical resolution is degenerated, the whole resolution of the picture intelligence of reproduction can not reach the degree that so degenerates yet.From the viewpoint of naked eyes sensation, as known in prior art, the vertical resolution of horizontal low frequency component is very important.
Figure 25 represents the embodiment of an improved circuit arrangement, is used for eliminating the non-uniformity that moves at the picture intelligence that reappears.In Figure 25, reference number 494 expressions are top with reference to the described the sort of decoder of Figure 18; 496 is a delay line; 498 is the low pass filters that the frequency component that allows to be lower than 4MHz is passed through; 500 is another delay line; 502 is the sub-sampling translation circuit; 504 is a phase place coincident circuit, and being used for sample frequency is that phase place and the sample frequency of the signal of 16MHz is that the phase place of the signal of 48MHz is consistent; 506 is an adder; 508 is a subtracter.
The MUSEII transmission signals not only is added to delay line 496, also is added to decoder 494.Decoder 494 decoding MUSE-II transmission signals, and an output signal, just the picture intelligence of Chong Xianing is added on the phase place coincident circuit 504.The MUSE-II signal that obtains from delay line 496 is added to sub-sampling translation circuit 502, so that the phase place of each line is met.The output of circuit 502 is added to "+" input of subtracter 508.The output of phase place coincident circuit 504 is added to "-" input of subtracter 508.
Phase place coincident circuit 504 is the sample frequencys from sub-sampling translation circuit 502 signal of 16MHz, adjust to sample frequency from decoder 494 be on the same phase place of the signal of 48MHz.Delay line 496 postpones to go the input signal of sub-sampling translation circuit 502, so that one of the signal delay of sub-sampling translation circuit 502 outputs is equaled the 504 output delay of output signal times of phase place coincident circuit.
The output of subtracter 508 is added to one of them input of adder 506 by low pass filter 498, and the output signal of decoder 494 is added to another input of adder 506 by delay line 500.Delay line 500 is used for making two input signals that are added to adder 506 to have same phase place.
Low frequency (lower than the 4MHz) component signal that does not have decoding of MUSE-II transmission signals has been carried out equal variation according to the theorem of well-known Tellegen.In the output signal of adder 506, a part of obscuring the reproduction picture intelligence that low frequency takes from decoder 594 has been eliminated by the output signal of low pass filter 498.
For example, can design as shown in figure 26 structure to low pass filter 498.In Figure 26, reference number 518,520,522,524,526 and 528 expression delay circuits, number number of their the power time picture elements that (4 ,-2 ,-1) expression is delayed.Reference number 530,532,534,536,538 and 540 expression adders; 542,544,548,550 and 552 is 1/2 multiplier.
The sort of low pass filter shown in Figure 26 does not need coefficient circuit, so its structure is very simple.In addition, as known in prior art, 1/2 multiplier can only be carried out 1/2 multiplication by displacement, also makes it structurally very simple.
Figure 27 shows the embodiment of a circuit, and it had both embodied the function of inserting filter 414 between the field, has also embodied the function of frequency translation/hybrid circuit 420.As shown in figure 27, seeing Figure 18 from blender 412() sample frequency that obtains is that the signal of 32MHz adds show up memory 416 and frequency changer 556 through input 554.
The signal that has postponed a field by field memory 416 is added to one of two inputs of adder 558, and after the signal that has postponed further postponed a horizontal line by one one row memory 560, is added to another input of adder 558.The output signal that obtains from adder 558 is added to sampling frequency conversion device 562.
Two frequency changers 556 and 562 clock rate from 2f
0(f
0=16.2MHz) transform to 3f
0Is the sample frequency that obtains from frequency changer 556 3f
0Output signal be added to and adopt switch 564 again, and be a sample frequency that obtains from sampling frequency conversion device 562 3f
0Output signal be added to another and adopt switch 566 again.Adopt switch 564 and 566 again for two input signal is modulated to 3f
0" 1 " of speed or " 0 " signal are (that is, with 3f
0Frequency input signal is squeezed into signal open and turn-off).By frequency is (3/2) f
0The clock signal of adopting again control and adopt switch 564 again, and be (3/2) f by frequency from inverter 568
0The clock signal of adopting again control and adopt switch 566 again.
From at the signal of adopting switch 564 output again, by a passband 3f
0/ 4(12MHz) logical filter 570 is added to multiplier 574.And from adopting the signal of switch 566 output again, be 3f/4(12MHz by a passband) high pass filter, be added to multiplier 576.In adder 578, the output signal addition of multiplier 574 and 576.Added result is added to frequency translation/blender 420(Figure 18 by an output 580).
According to the signal that is added to multiplier 574 and 576, determine the mixing ratio between the output signal of the output signal of low pass filter 570 and high pass filter 572, so that be consistent with visual motion from motion detector (seeing Figure 18) output.
Figure 28 shows the embodiment of another circuit arrangement, and it had both embodied the function of inserting filter 414 between the field, has also embodied the function of frequency translation/blender 420.Figure 29 represents the specific embodiments of a variable coefficient filter shown in Figure 28.
With reference to Figure 25, seeing Figure 18 from blender 412() the sample frequency of broadcasting of output is that the signal of 32MHz not only has been added to a memory 416, also has been added to variable coefficient filter 582.The output signal of adder 558 has been added to another variable coefficient filter 584.
Refer now to Figure 29, variable coefficient filter 582(584) have two switches 586 and 588, and one group of coefficient circuit 590.The clock signal of adopting again by (3/2) f is come control switch 586, and sees Figure 18 according to motion detector 408() output signal come control switch 588.
Generally, can be a read-only memory (ROM) as the variable coefficient filter.If the output signal of signal and detector 408 can both be ended input as seeking of this ROM during the sampling of 3/2f, then only just can realize this variable coefficient filter with a string ROM.Draw thus, can focus on one group of one dimension variable coefficient filter in the bidimensional variable coefficient filter.
As mentioned above, can realize adopting again function and insertion by a string variable coefficient filter.Therefore, can be the compact in size of inserting filter between the field, and do not reduce its performance.
According to the present invention, when between detect adopting, compiling the motion parts that moves in the transmission system that sampling and interframe skew sample, can utilize frame difference signal, thereby promote accuracy of detection and it is simplified the structure.
And the present invention has an advantage, and Here it is, even the wrong detection of motion parts has taken place, also can not cause degenerating of any essence on image quality.
In addition, according to the present invention, can accurately detect the motion of image in MUSE-II signal transmission, thereby improve the quality of reappearing image.
According to the present invention, can eliminate the non-uniformity of the picture motion that is caused by the decoder that reappears picture intelligence and noise decrease, thereby promote visual stability.
In addition, according to the present invention, can eliminate the visual constant motion that the compensation by the picture motion of non-complete motion vector causes.
And, according to the present invention, can be reduced in desired C/N in the transmission line.
According to the present invention, can reappear a picture intelligence with fine vertical resolution.
At last, according to the present invention, can be the compact in size of inserting wave filter between the field, and can not cause any reduction of its performance.
Claims (4)
1, a kind of transmission system that is used for (MUSE) bandwidth reduction of improved multichannel Nyquist sampling coding, skew and sampling and interframe skew and sampling are compressed the encoder that also detects movable information with defeated this TV signal of Fu simultaneously to a bands of television signals between the employing field, it is characterized in that above-mentioned encoder comprises.
First sampler (46).This device carries out being offset subsampling between the field to vision signal in the above-mentioned TV signal according to the timing of sampling pulse, and this sampling pulse has and above-mentioned vision signal and corresponding basically first sampling frequency of upper limiting frequency (24.3MHZ).Obscure part and have by what skew subsampling between above-mentioned produced from the output signal of above-mentioned first sampler, output signal waits on half the frequency of above-mentioned first sampling frequency (12.15MHZ) be confused (seeing 16 pages (3)) substantially in frequency;
Second sampler (52), this device carries out being offset subsampling between the field to the above-mentioned vision signal in the above-mentioned TV signal according to the timing of an above-mentioned sampling pulse, and above-mentioned sampling pulse has first sampling frequency (24.3MHZ) (see 16 pages (6));
Motion detection apparatus (56), this device detect corresponding to the motion in the image of above-mentioned vision signal: (seeing 16 pages (8));
Mixer device (58), this device are used for according to from the output signal of above-mentioned motion detection apparatus (56) output signal from first and second samplers being carried out mixing: (seeing 16 pages (9))
The 3rd sampler (60), this device receives the output signal from above-mentioned mixer device (58), be used for carrying out interframe skew subsampling by the timing of a sampling pulse, this sampling pulse has one second sampling frequency (16.2MHZ), this sampling frequency is lower than above-mentioned first sampling frequency (24.3MHZ), but be higher than half of above-mentioned first sampling frequency, from above-mentioned interframe skew subsampling, produced a multichannel subsampling transmission signals partly of obscuring that does not comprise paraphase like this.(seeing 16 pages (10))
2, encoder as claimed in claim 1 is characterized in that above-mentioned first sampling frequency of carrying out between the field skew subsampling and the above-mentioned ratio of carrying out interframe and being offset between second sampling frequency of subsampling are decided to be 3: 2 in the ranks.
3, the bandwidth reduction transmission system of a kind of improved multichannel Nyquist sampling coding (MUSE), this system comprise one adopt between skew subsampling and interframe skew subsampling one bands of television signals is compressed to transmit this TV signal, and detect the encoder of movable information simultaneously, it is characterized in that above-mentioned encoder comprises:
Prefilter between one (154), this filter receives an incoming video signal, and be used for transmitting a signal corresponding to the bandwidth of first transmission characteristic, this characteristic the ceiling restriction by the horizontal component frequency of prefilter between above-mentioned field to first sampling frequency (24.3MHZ);
First sampler (156), this device receives the output signal of prefilter between above-mentioned, and according to the timing of sampling pulse, above-mentioned output signal is carried out being offset subsampling between the field with first sampling frequency, this sampling pulse have above-mentioned first sampling frequency (24.3MHZ);
Low pass filter (158), this filter receives an output signal from above-mentioned first sampler (156), and make have obscure the part signal component pass through, this obscures part by producing in the skew subsampling between above-mentioned field, and this low pass filter also has first cut-off frequency (12MHZ) that a reality equates with above-mentioned first sampling frequency (24.3MHZ) half (1/2);
First converter apparatus (160), this device receives the output signal of above-mentioned low pass filter (158), and the sampling frequency (48.6MHZ) of above-mentioned output signal is converted to a different sampling frequency (16.2MHZ);
Prefilter (162) between, this filter receives above-mentioned incoming video signal, and be used for transmitting its bandwidth and the corresponding signal of second transmission characteristic, this characteristic the ceiling restriction by the frequency of the horizontal component of prefilter between above-mentioned field to second cut-off frequency (12MHZ);
Second converter apparatus (164), this device receives the output signal of prefilter between above-mentioned, and the sampling frequency (48.6MHZ) of above-mentioned output signal is transformed to a different sampling frequency (32.4MHZ);
Motion detection apparatus (166), the above-mentioned incoming video signal of this device response detects the motion of an images;
Mixing arrangement (168), this device mixes the output signal of above-mentioned first and second converter apparatus (160,164) according to the output signal of above-mentioned motion detection apparatus; And
Second sampler (170), this device receives the output signal of above-mentioned mixing arrangement (168), and, according to the timing of a sampling pulse above-mentioned output signal being carried out interframe skew subsampling, this sampling pulse has one second sampling frequency (16.2MHZ).This frequency is lower than above-mentioned first sampling frequency (24.3MHZ), but is higher than above-mentioned sampling frequency half (1/2), like this, does not comprise that the second sampling transmission signals of obscuring multichannel partly of paraphase just can be produced by above-mentioned interframe skew subsampling.
4, encoder as claimed in claim 3 is characterized in that carrying out first sampling frequency of above-mentioned skew secondary sample and the ratio of carrying out above-mentioned interframe and being offset second sampling frequency of subsampling in the ranks is decided to be 3: 2.
Applications Claiming Priority (7)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP106130/85 | 1985-05-20 | ||
| JP106132/85 | 1985-05-20 | ||
| JP60106132A JPS61264889A (en) | 1985-05-20 | 1985-05-20 | Transmission system for multiple subsample |
| JP13,632/86 | 1986-01-27 | ||
| JP13,631/86 | 1986-01-27 | ||
| JP61013632A JPS62172876A (en) | 1986-01-27 | 1986-01-27 | Motion detecting system for multiplex subsample transmission signal |
| CN86103338.8A CN1009898B (en) | 1985-05-20 | 1986-05-19 | Bandwidth compressed transmission system |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN86103338.8A Division CN1009898B (en) | 1985-05-20 | 1986-05-19 | Bandwidth compressed transmission system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1041502A true CN1041502A (en) | 1990-04-18 |
| CN1013016B CN1013016B (en) | 1991-06-26 |
Family
ID=27178846
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 89101690 Expired CN1013015B (en) | 1985-05-20 | 1986-05-19 | Decoder of bandwidth compressed transmission system |
| CN 89101692 Expired CN1013016B (en) | 1985-05-20 | 1986-05-19 | Coder of bandwidth compressed transmission system |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 89101690 Expired CN1013015B (en) | 1985-05-20 | 1986-05-19 | Decoder of bandwidth compressed transmission system |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN1013015B (en) |
-
1986
- 1986-05-19 CN CN 89101690 patent/CN1013015B/en not_active Expired
- 1986-05-19 CN CN 89101692 patent/CN1013016B/en not_active Expired
Also Published As
| Publication number | Publication date |
|---|---|
| CN1013015B (en) | 1991-06-26 |
| CN1013016B (en) | 1991-06-26 |
| CN1041501A (en) | 1990-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN86103338A (en) | The transmission system of bandwidth reduction | |
| CN1036430C (en) | Wide screen television | |
| CN1048131C (en) | Digital video camera apparatus | |
| CN1112807C (en) | Partial response trellis decoder | |
| CN1178472C (en) | Image signal system converter and TV set | |
| CN1167006C (en) | Television receiver, video signal processing device, image procesing device and image processing method | |
| CN1064206C (en) | Image processing apparatus | |
| CN1297929C (en) | Data processing equipment and data processing method and the recording medium | |
| US5132793A (en) | Television receiver compatible with both standard system television signal and high definition television signal | |
| CN1134975C (en) | Image processing device | |
| CN1039472C (en) | Apparatus for detecting position of area of words screen and picture charactor such as pictureless zone | |
| CN1170317A (en) | Process for interpolating progressive frames | |
| CN1237067A (en) | Picture encoding and decoding apparatus, method and data storage medium | |
| CN1018136B (en) | Television receiver with static stores switch signal | |
| CN1249882A (en) | Video signal reproducing device and hierarchical video signal decoder | |
| CN1158863C (en) | Video signal reproducing device | |
| CN1013016B (en) | Coder of bandwidth compressed transmission system | |
| CN1044197A (en) | Handle the method and apparatus of TV signal | |
| CN1179571C (en) | Apparatus and method for mapping and image to blocks to provide for robust error relovery in lossy transmission environmental | |
| CN1173552C (en) | Video display apparatus with vertical scan velocity modulation and video display method therefor | |
| CN1075587A (en) | With TV-related improvement | |
| CN1510908A (en) | Solid camera device, driving method thereof and camera | |
| CN100342734C (en) | Video signal processing circuit and video signal processing method | |
| CN1076577A (en) | The improvement of television system | |
| CN1240207C (en) | Digital video camera apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C13 | Decision | ||
| GR02 | Examined patent application | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |