DE4405351A1 - Video motion detector - Google Patents

Abstract

A video motion detector for colour video signals operates on the chrominance component signals by comparing the signals on the current frame with those on the previous frame (20, 22). This method can incorrectly indicate chrominance motion when in fact there is no chrominance in the current frame e.g. when uncoloured luminance detail is uncovered by a moving coloured object. To overcome this problem, a rectifying and clipping circuit (32) determines whether there is chrominance in the current frame, and, in the absence of chrominance in the current frame, forces the motion output signal to indicate the absence of motion. The motion detector is particularly useful with the PALplus broadcast television proposal. <IMAGE>

Description

The invention relates to a video motion detector for generating a motion output signal that indicates whether there is motion in a color video signal. The Be Motion detector is for use in a motion adaptation suitable video processing system.

The invention has been developed for use in connection with the proposed PALplus widescreen television system, which was developed by A. Ebner et al. in the publication "PALplus - The European System for Wide-Screen Terrestrial TV", International Broadcasting Convention, Amsterdam, 3-7 July 1992, IEE Conference Publication 358, pages 203-207. In this system, an image with an aspect ratio of 16: 9 is transmitted so that it can be reproduced by an improved receiver with an aspect ratio of 16: 9, but can also be reproduced compatible by a conventional receiver with an aspect ratio of 4: 3 . This is achieved in that the 16: 9 aspect ratio signal is only transmitted on three quarters of the available lines. The middle three-quarters of the picture, which are also called "mailboxes", are selected for this purpose, black stripes (bars) remaining on the upper and lower edge of a conventional recipient. Auxiliary signals of low amplitude are transmitted in these black stripes and are used in the improved receiver to improve the quality of the image derived from the mailbox lines. They are not visible on the screen of a conventional receiver.

Complicated processing for generating takes place in the transmitter transmission signals and also in the improved receiver to form signals for the implementation of the transmission signal into a pictorial representation. In the mentioned publication it is proposed to use different processing methods in Depending on whether the video signal originally was derived from a camera or a film. If it was derived from a camera, they provide the 20ms fields of each 40ms frame (or picture) true situation at 20 ms apart times. However, this is not the case when the video signal is off was derived from a film because the two were nested Fields derived from a single frame of a film are. They therefore provide information of the same Time. A better picture can be displayed under Taking this fact into account and using ver various filter functions both in the transmitter and in every improved receiver in camera and film Achieve operation.

It is understood that in a similar way a better picture representation can be achieved if a motion detector is used and set up in camera mode works to identify areas where no movement in full screen between adjacent fields. These areas can then be analogous to those in film mode processed, making use of the fact is that the two fields of the same image information correspond. This requires a motion detector in the Sen and a corresponding motion detector in each ver better recipient.

In particular, the document mentioned proposes to apply a coding method based on that of R. Kays in the essay "A Process for Improved PAL coding and decoding "in the publication" Fern seh- und Kinotechnik ", Volume 44, No. 11, November 1990, Pages 595-602 based, which is called Color-Plus  System is called. This works with line pairs, which are 312 video lines apart, d. H. with a Line from each field of the full screen. The above of around 3 MHz lying luminance component is in the two Matched rows in a pair, such as the Mean is formed from both, and in a similar manner the chrominance components in the two lines are one Couple adjusted to each other. This results in the decoder Forming the mean and difference of the rows of each pair a neatly separated luminance and chrominance.

The method proposed as PALplus is an adaptive one Variant of Color-Plus, in which between the Color-Plus-Be driven and a fall-back mode under the Control of a motion detector is switched. The Synchronization of the switching processes in the encoder and decoding rer is achieved by using the same motion detector circuit in the transmitter (encoder) and in the improved receiver (Decoder) reached. Because of the Bil just described the average of the two by 312 lines apart of the lines in the picture, is the vertical temporal sliver te of chrominance and luminance above 3 MHz limited. For areas with color movement is therefore preferred suggest that no luminance is transmitted above 3 MHz and instead of averaging two lines apart by 312 image lines only the unmodified modulated chrominance in this part of the spectrum are transmitted.

However, it was found that when the end of movement tector is malfunctioning and a movement is detected where there is actually no high-frequency Lumi nanzdetail unnecessarily blurred (muffled or abge weakens).

The invention is defined in claim 1. Training the invention are specified in the subclaims.  

In a preferred video motion detector, the next one Input Chromi are described in greater detail nance component signals of the current frame with compared to the previous frame. As before may incorrectly result in chrominance evidence be displayed, although actually at the moment full frame no chrominance is available. To the solution This problem is checked whether in the current full picture Chrominance is present. If there is no chroma in the au the current full screen is available, the motion Output signal inevitably set to a value that represents the absence or absence of movement.

The invention is explained in more detail below with the aid of a Embodiment with reference to the drawings described. Show it:

Fig. 1 is a block diagram of a previously proposed Be wegungsdetektorschaltung,

Fig. 2 shows a corresponding block diagram of a contemporary movement detector circuit Invention,

Fig. 3 is a block diagram of an improved movement detector circuit, which is based on the circuit of FIG. 2, and

FIGS. 4 and 5 are block diagrams of an encoder and a decoder in which the movement detector may be used.

The previously proposed motion detector circuit 10 shown in FIG. 1 processes an intra-frame averaged chrominance, that is, the average of each pair of lines that are 312 video line periods apart to produce a motion signal M. The motion signal M should be low in areas in which the chrominance is stationary or non-existent and high in areas in which it is moving.

For this purpose, the motion detector circuit 10 receives a baseband chrominance component signal U and V at inputs 12 and 12 A. In the two channels, similar processing takes place, so that only the U channel is described in detail. The corresponding components in the V-channel are identified by the same reference numbers and the index A.

The signal U at the input 12 is a horizontal low pass filter 14 with the filter function HF1 and from there on the one hand via a one-line delay element 16 and on the other hand di one of the inputs of an adder 18 leads directly. The adder practically effects the formation of the mean value from the two input signals supplied to it (the required amplitude halving being taken into account at any suitable point in the circuit). The output signal of the adder 18 is then supplied on the one hand via a frame delay element 20 and on the other hand each directly to one of the inputs of a subtractor 22 , so that the signal from the previous frame is subtracted from the signal for the current frame in order to Form differential signal. The frame difference signal is supplied to a rectification and clipping circuit 24 . This can be formed by a first look-up table (LUT), usually in the form of a read-only memory (ROM). The U-output signal of the circuit 24 is then fed together with the corresponding V-output signal of the circuit 24 A to a functional circuit 26 , which can also be a look-up table, again in the form of a ROM, and which outputs the motion-indicating output signal M forms.

. During operation of the motion detector circuit 10 of Figure 1 is - after the horizontal low-pass filtering with a cutoff frequency of about 1 MHz in the filter 14 - averages the chrominance in the adder 18 via the line delay 16 ge, any one to the influence of static phase error in the received signal as to prevent with normal PAL decoding by means of a delay line. Then the frame delay 20 is used to calculate the amplitude difference between the signals of the current frame and the previous frame. A high amplitude difference signal at the output of subtractor 22 means the presence of a chrominance movement. This difference signal is then rectified and trimmed (cut off) by the look-up table LUT1 in the circuit 24 .

The motion output signal M is then determined by the functional circuit 26 from the U and V signals by calculating the size of the equivalent PAL chrominance signal. Theoretically, this is represented by the equation of an ellipse in UV coordinates:

M = [(0.886U / 2.02) ² + (0.701V / 1.14) ² ] ^1/2 .

In practice, a simpler expression can be used the, e.g. B. a sectionally linear approximation to the El lipse.

It has been found that the circuit of Fig. 1 has a significant disadvantage. It correctly generates a high value of M if there is chrominance in the current frame but none is present in the previous frame. However, it also produces a high value of M if there was chrominance in the previous frame, but none in the current frame. In the overall system, this appears as a damping (soft focus) of a largely colorless luminance detail when it is exposed by a moving colored object.

This problem is solved by the motion detector circuit 30 according to the invention shown in FIG. 2. This largely coincides with the motion detector 10 of FIG. 1, and the corresponding components are provided with the same reference numerals and will not be described again in detail below.

The additional components in FIG. 2 comprise a further rectification and cut-off circuit 32 , which is connected to the output of the adder (averager) 18 . The output signal of the circuit 32 together with the output signal of the corresponding circuit 32 A is fed to the V-channel of a functional circuit 34 . Circuits 32 and 34 are in turn formed as look-up tables and are similar to circuits 24 and 26 . The output signal of the functional circuit 34 is fed to a further look-up table 36 which has a threshold function. A multiplier 38 receives the output signal of the main function circuit 26 and modifies it by multiplication by the output signal of the circuit 36 to form the motion signal M at the output of the motion detector.

In the circuit of Fig. 2, the motion signal of the functional circuit is therefore multiplied by 26 to the output signal of the threshold circuit 36, whereby a control signal in dependence on the amplitude of its input signal is formed. The output of the threshold circuit 36 is zero when there is no chrominance in the current frame, blocking the wrong motion signal generated by the circuit of FIG. 1. The output of the threshold circuit 36 is one when there is significant chrominance in the current frame so that it allows the passage of the correct motion signal through the multiplier 38 to the output.

Intermediate values can be generated at lower chrominance levels become; this can help avoid abrupt and visible Operating mode changes can be helpful.  

While a frame delay element is used in FIG. 2, in principle a delay element with a field delay, possibly with a corresponding interpolation, could be used instead.

As has been shown, the circuit of FIG. 2 has another advantage. This is particularly the case when a colored object moves against a complementary colored background. The chrominance component at the edge of the object goes through zero, and because of the blurring effect of the horizontal low-pass filter 14 , 14 a, the chrominance at this edge can be small over a considerable number of pixels. In this area, the circuit of Fig. 2 may erroneously produce a low value of M since there is obviously no chrominance in the current frame. When this occurs, chrominance is transmitted with reduced vertical-temporal bandwidth. This can result in 25 Hz flickering in horizontally moving narrow colored lines. In the circuit of Fig. 1, this problem does not occur because the edge is moving, so that a high difference signal occurs.

This further problem is solved by the motion detector circuit 50 of FIG. 3, in which the horizontal filters 14 , 14 A lying in front of the line delay elements 16 , 16 A are omitted and instead four separate horizontal low-pass filters 52 , 52 A; 54 , 54 A are provided. The filter 52 is arranged between the rectification and cut-off circuit 24 and the functional circuit 26 and the filter 54 between the rectification and cut-off circuit 32 and the functional circuit 34 . This arrangement of the horizontal low-pass filter behind the rectification and clipping circuits maintains the low amplitude chrominance at the edge of the object so that the correct value of M is formed.

In Fig. 3, many of the processing blocks are identical. These are shown connected by vertical dashed lines, and the circuitry can be reduced by using only one component instead of the blocks connected by the dashed lines and multiplexing the signals passed through them.

The function of the multiplier 38 can be performed by the function circuit 26 , which would then have three inputs, or it can be performed by the lookup table 36 , which would then have two inputs. Another modification, which is not shown, can be that the two threshold value functional components 36 and 38 are replaced by a comparator. In this embodiment, a look-up table receives the outputs of circuits 24 and 32 A and generates a signal M ₁ that represents the chrominance movement, while another look-up table receives the outputs of circuits 24 A and 32 to produce a signal M ₀ that corresponds to the Represents the amount of chrominance in the current full screen. The comparator receives the signals M ₁ and M ₀ and selects the smaller of the two input signals to form the movement output signal M. This is then converted into corresponding signals for the respective control of the luminance and chrominance levels using further look-up tables.

FIG. 4 shows a block diagram of the essential components of an adaptive color plus encoder 60 , which has a motion detector 50 of the type shown in FIG. 3 (or in FIG. 2). The circuit is simplified by omitting the delay elements and filters, which are not necessary to understand the system, but the mode of operation is clear to the person skilled in the art.

Inputs 62 , 64 , 66 receive components Y, U and V of an interlaced television signal (an interlaced television signal). Via (not shown) delay circuit and switches, line pairs from each frame are fed to the encoder simultaneously, one line from each field. The phase position of this line pairing is important because the Color Plus process is based on the phase position of the color subcarrier being reversed after 312 line periods. During the first output field, the output line n is encoded from the input lines n and (n + 312), e.g. B. the output line 24 from the input lines 24 and 336 . During the second output field, the output line n is made up of the input lines (n-312) and n, e.g. B. the output line 336 from the input lines 24 and 336 , coded.

The operation of the encoder is as follows. The luminance input signal Y is divided by the horizontal filter 68 with the function HF2 and the connected subtractor 70 into frequencies above and below (approximately) 3 MHz. The low frequencies are switched directly to the PAL output via a field repetition frequency switch 72 . This switch is shown in the "field 2" position. The high frequencies are averaged by an adder 74 (which simultaneously divides by 2) before they are added back to the low frequency luminance in the adders 76 . The amplitude of this portion of the signal is controlled by the motion detector: when there is moving chrominance, the "Y control signal" drops from one to zero, thereby blocking the high-frequency luminance by the action of multiplier 78 . Overall, this ensures that the luminance above 3 MHz in the PAL output signal is the same in every row of a pair, but is not present if there is a moving chrominance.

Chrominance signals U and V are averaged and subtracted by adders 80 and subtractors 82 . The intra-frame average signals are supplied to the motion detector circuit 50 and are also passed through the PAL output via the field repetition frequency switch 84 and subcarrier frequency fsc modulators 86 . If there is moving chrominance, the "UV control signal" from the motion detector rises from zero to one so that the intra-frame differential chrominance can be added to the output signal due to the action of multipliers 88 . This ensures overall that the unmodulated chrominance signals are the same in every row of a pair, except that the chrominance moves; in this case the entire chrominance signal is sent. The modified intra-frame difference signal and the mean signal are added or subtracted in adders 90 and subtractors 92 before they are supplied to the selector switches 84 . The outputs of the modulators 86 are added in the adder 94 , the output of which is added to the luminance signal by the adder 96 to form the PAL output.

The corresponding decoder 100 is shown in FIG. 5. The PAL input signal at inputs 102 is converted into line pairs, one from each field, as in the encoder of FIG. 4. Identical band-splitting filters 104 with the filter function HF2 are used to determine a low-frequency luminance which are supplied to the luminance output via adders 106 and a field repetition frequency switch 108 . The low frequency components are subtracted from the input signals in subtractors 110 to form a high frequency luminance plus modulated chrominance. This is averaged and subtracted by an adder 112 and a subtractor 114 .

Since the color subcarriers have opposite phase positions in each line of an input pair, the intra-frame average signal has no remaining subcarrier if there is no moving chrominance. This signal therefore represents the pure high-frequency luminance and is added to the low-frequency luminance by the adders 106 after it has been modified by a multiplier 116 in response to a "Y control signal" from the motion detector.

Similarly, the intra-frame difference signal contains modulated chrominance with no residual luminance. Demodulators 118 , 120 , 122 and 124 demodulate the output signals of the adder and subtractor 114 with respect to the U and V phase positions, respectively. After demodulation, the intra-frame difference signal forms the pure U and V-intra-frame mean signals which have been coded. These are fed to the motion detector 50 and further via the adders 126 , subtractors 128 and field repetition frequency switches 130 to the outputs U and V.

If there is chrominance movement, the entire PAL signal above 3 MHz is used for chrominance to form the full vertical temporal bandwidth. In this case, the radio frequency luminance signal is turned off by multiplier 116 because the "Y control signal" drops to zero. At the same time, the intra-frame difference signal is added by multipliers 132 to the chrominance output signals since the "UV control signal" rises to one.

If the motion detector of Fig. 2 or Fig. Is used in the circuit of Fig. 4 3, the look-up table 34 has a higher by 3 dB gain as the look-up Table 24. The reason for this is that the intra-frame difference channel (M ₁ ) introduces noise from both frames, while the current frame channel (M ₀ ) only introduces noise from one frame.

Generally, the system could go through in whole or in part Software can be realized. In this case they would be To consider figures as flow diagrams or logic diagrams.

It should be noted that the execution described example of a 625/50 interlace PAL color video signal applies, but the principle described also applies to other types of color video signals is applicable.

Claims (7)

A video motion detector comprising: an input ( 12 ) for a chrominance component signal; one connected to the input delay means for delaying the chrominance component signal by a video frame or field and comparison means ( 22 , 24 , 26 ) connected to the input and output of the delay means to the chrominance component for the current frame with that compare for the previous frame or field and produce a motion output signal having a first value when the differences between the current and previous frame or field chrominance components are relatively small, indicating the absence of motion, and has a second value if these differences are large, indicating the presence of motion, characterized in that chrominance detection means ( 32 , 34 , 36 , 38 ) is connected to the input to determine whether the current frame or field a chrominance exists and to set the motion output to the first value when there is no chrominance in the current frame or field. 2. Video motion detector according to claim 1, characterized in that a horizontal low-pass filter ( 14 ; 52 ) with the delay means is connected in series. 3. Video motion detector according to claim 2, characterized in that the horizontal low-pass filter ( 52 ) is connected in series with the output from the delay means and the Chromi nanzdetektionsmittel has a second corresponding horizontal low-pass filter ( 54 ). 4. Video motion detector according to one of the preceding claims for interlaced video signal, characterized in that it has a means ( 16 , 18 ) for forming the mean value of adjacent lines of the input chrominance component signal. 5. Video motion detector according to one of the preceding claims, characterized in that it has separate inputs ( 12 , 12 A) and delay means ( 20 , 20 A) for two chrominance component signals and the comparison means and the chrominance detection means which process the chrominance component signals. 6. Video transmitter with a motion detector according to one of the preceding claims. 7. Video receiver with a motion detector after one of the preceding claims.