JPH0569352B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a multi-subsampled transmission signal, in particular,
A multiple sub-sampled transmission signal that does not include aliasing components between frames in the low frequency range (this transmission method was
The present invention relates to a motion sensitivity control method in a decoder (described in detail in No. 106,132).
Further, the present invention can be applied to a system in which a motion signal is obtained from an image signal subsampled in frames or fields.

[Summary of the Disclosure] One of the methods for band compression of television signals is a multiple sub-sample transmission method using inter-frame and inter-field offset subsampling, such as MUSE (Multiple Sub-Nyquist).
There is a current high-definition television signal multiplex sub-sampling transmission method called Sampling Encoding, which effectively achieves band compression. The details are described in, for example, the Institute of Electronics and Communication Engineers Technical Research Report Image Engineering IE84-72.

In such a decoder for multiple sub-sampled transmission signals, the final image is obtained by switching between a still image and a moving image, and the spread of the motion signal that performs this control must be equal to the moving area. The motion signal is obtained by taking the difference between frames and passing it through a low-pass filter. In order to obtain an appropriate motion signal, it is necessary to perform the following two processes on the output signal of the low-pass filter.

1. Processing for not outputting a motion signal in a still area 2. Processing for outputting a motion signal in a moving area The motion signal obtained from the low-pass filter output may be partially omitted in the moving area, or if there is movement in the still area. However, such errors are likely to occur at the boundary between the moving area and the stationary area.

Processing 1 above is a process for narrowing the region of the motion signal, and processing 2 above is processing for widening the region of the motion signal. Therefore, these two processes are contradictory processes, and their methods are also different.

The present invention relates to the above item 1, and uses vertical correlation of images to narrow the moving area and prevent the moving area from expanding to a portion that is originally a still area. The present invention prevents still images from being erroneously reproduced as moving images,
Reduce unnecessary aliasing signals from output images.

[Prior art] In the reproduction of multiplex sub-sampled transmission signals, there are examples in which inter-frame differences and edges are combined to obtain a motion signal (if the inter-frame differences are simply used as a motion signal, the sensitivity is high at the edge portions of the image). (The sensitivity is adjusted by dividing the inter-frame difference by the edge detection amount.) However, there is no example of adding vertical correlation to this.

[Problems to be Solved by the Invention] In the conventional motion signal detection method in the reproduction method of multiple subsampled transmission signals, the motion signal spreads too much, especially when there is a thin vertical line in the screen. was determined to be a partially moving area. As a result, there are regions in which vertical lines are reproduced as moving parts and regions in which they are reproduced as static parts. In the image reproduced in this way, the vertical lines become discontinuous. In actual images, edges such as the edges of buildings are affected by this and disappear.

[Means for Solving the Problems] The present invention is characterized in that a motion signal is controlled by applying vertical correlation in multiple sub-sampled transmission signal reproduction.

[Embodiment] Since the multiple subsample transmission method uses subsamples, the motion signal is also obtained from the subsampled signal. A specific example of the problem when a motion signal is obtained from a subsampled signal will be described below.

The image shown in Figure 1 is a case where there is a black vertical line on the screen, and the single circle in the figure is the current frame.
Double circles indicate sampling points of the previous frame.
The method of detecting the motion signal of the point is obtained by |1/4(〓+〓)-1/2|.

In this case, black is black, and 〓 and 〓 are white, so the motion signal of the point has a sufficiently large value. The points for this purpose are determined to be dynamic areas, and interpolated using the four points 〓, 〓, 〓, 〓. However, although this point is originally a stationary area, it has become a moving area because a motion signal was created from the sub-sample data, so the motion signal at this point should not originally be output.

If the vertical high-pass filter is configured in the form |1/4(〓+〓)-1/2|, then , 〓, and 〓 are all black, so their value becomes zero. The motion signal can be controlled using the output of this filter. That is, when the filter output is low, the value of the motion signal may be reduced.

The configuration of such a vertical high-pass filter (abbreviated as HPF) is shown in FIG.

In Fig. 2, 1 and 2 are 1H (horizontal period) delay circuits, 3 and 4 are adders, 5 is a multiplier with a multiplier of 1/2,
6 is an exclusive OR gate, 7 is an exclusive OR gate with an inverted input, and 8 is an absolute circuit.

There are two possible methods of suppressing a motion signal using the HPF output of such a configuration.

1 Method of suppressing motion signals using the HPF output value 2 Method of suppressing motion signals by changing the edge amount using the HPF output Two methods are used for transmission methods that do not include aliasing components between frames in the low frequency range of the transmission signal. ing.

The motion signal is obtained by processing the frame difference with a low-pass filter and then dividing by the edges, but in this method, the amount of edges is controlled by the HPF output.

That is, when the HPF output is small, the amount of edges is increased, and the low-pass filter output of the frame difference is divided by a large value. As a result, the value of the motion signal becomes smaller and the motion area becomes smaller.

Figure 3 shows a system diagram of this method. In FIG. 3, 9 is a low pass filter, 10 is an edge table, and 11 is a divider.

As mentioned in the explanation up to this point, this method uses conventional edges (edges in Figure 3) and vertical edges.
The edge amount is newly determined based on the HPF output. (In Figure 3, this newly determined edge amount is indicated as edge).

FIG. 4 is a three-dimensional representation of the relationships among these three edges, edges, and HPF outputs, and the edge table 10 in FIG. 3 has the characteristics shown in FIG. 4.

In a transmission system that does not include inter-frame aliasing components in the low frequency range of the transmission signal, a method of controlling the edge amount using the HPF output is adopted as described above, but the motion signal may also be directly controlled using the HPF output.

[Effects of the Invention] According to the present invention, an appropriate motion signal can be obtained using simple hardware. This makes it possible to improve the reproduced image quality.

[Brief explanation of the drawing]

Fig. 1 is a principle diagram of motion signal detection in a sub-sample image, Fig. 2 is a circuit diagram of a vertical high-pass filter, Fig. 3 is a system diagram of motion signal detection in the present invention, and Fig. 4 is a diagram of the present invention. FIG. 3 is a three-dimensional diagram for determining the value of an edge signal in the invention. 9...Low pass filter, 10...Edge table, 11...Divider.

Claims (1)

[Claims] 1. In the reproduction of multiple subsampled transmission signals of image signals, the level of the motion signal obtained from the inter-frame difference of the transmission signals is controlled to be reduced as the vertical correlation of the image signals increases. Features a motion sensitivity control method for reproducing multiple subsampled transmission signals. 2. A motion sensitivity control method for reproducing multiple sub-sampled transmission signals as set forth in claim 1, characterized in that the motion signal is controlled by an output of a vertical high-pass filter of the image. Motion sensitivity control method. 3. In the motion sensitivity control method for reproducing multiple sub-sampled transmission signals as set forth in claim 1, the motion signal is controlled by the output of a high-pass filter in the vertical direction of the image and the edge of the image. Motion sensitivity control method for transmission signal reproduction. 4. The motion sensitivity control method for reproducing multiple sub-sampled transmission signals as set forth in claim 1 or 2, characterized in that the vertical high-pass filter used for the control is configured by a 1H delay and an exclusive OR. A motion sensitivity control method for multi-subsampled transmission signal reproduction.