JPH0681299B2 - Image quality improvement device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture processing when a still picture processing and a moving picture processing are performed independently in a band-compressed television signal receiving apparatus. The present invention relates to a device for improving image quality.

2. Description of the Related Art When transmitting a wideband video signal such as a high-definition television signal, a method of band-compressing the video signal has been studied. One of them is MUSE (Multiple Subnyquist Samp).
ling Eucoding) method was proposed by the Japan Broadcasting Corporation.
The details of the contents are shown, for example, in Yuichi Ninomiya et al., "High-definition Television Satellite 1-Channel Transmission System (MUSE)", The Institute of Electronics and Communication Engineers, Technical Report, IE84-72, 1984.

In the MUSE method, the pixels of the video signal sampled at sHz are offset for each field and frame and thinned to 1/4. Is transmitted, and the receiving side reproduces the screen by superimposing the following four fields to obtain a high-quality image.
When processing a still image, the information in the past field may be used as described above, but in the case of a moving image, another processing is required.

In the case of a moving image, afterimages and the like are conspicuous when processing is performed using information of past fields as described above. Therefore, for a moving screen, the image is reproduced only by the received field.

Since the processing of this still image and moving image is too coarse if it is switched for each screen, in reality, a motion detection circuit that detects moving pixels is used to determine the moving part and the still part within the screen and detect that. Depending on the signal, the signal subjected to each processing is output.

However, there are some scenes in which sufficient image quality cannot be obtained by such a processing method alone. For example, when the entire screen slowly moves in the same direction by moving the camera horizontally or vertically, the moving image processing is performed, and the image blur becomes conspicuous. In order to solve this problem, a process called motion correction is performed. For the motion correction processing, a motion vector signal indicating the motion direction and motion speed of the entire screen is detected on the transmitting side and transmitted together with the video signal for each field. On the receiving side, the past video signal stored in the memory is read out from the memory by shifting it by the amount moved by the motion vector signal, and it is possible to handle it as a still image by overlapping the past and present image signals and prevent image quality deterioration. I'm out.

Hereinafter, the above configuration will be described with reference to the drawings. In FIG. 2, 1 is a frame memory without motion compensation (non-motion compensation type frame memory), 2 is frame memory with motion compensation (motion compensation type frame memory), 3
Is an input switching circuit for superposing a received input signal and a signal of one frame before and interpolating between frames, and 4 is a motion detection circuit (I) for determining a motion part and outputting a motion detection signal, 5 is a still image processing circuit, 6 is a moving image processing circuit, 7
Is a mix circuit for selecting or synthesizing a still picture and a moving picture according to the motion detection signal from the motion detection circuit (J), 4.

First, a case where motion correction is not performed will be described.
Now, consider the case where, as shown in FIG. 3a, the parts other than the round part move leftward at a constant speed, and only the round part is stationary. Assuming that the input image is the one shown in FIG. 3b, the outputs of the non-motion-correction type frame memory 1 and the motion-correction type frame memory 2 are image signals 1 frame and 2 frames before, respectively, and are c and d. The motion detection circuit (I) 4 detects the motion part by comparing the current frame and the past frame signal. For example, the motion detection circuit (I) 4 outputs a signal for comparing e and d and determining that the circular portion is a still image and the other shaded portion is a moving image like f. With this signal, the mix circuit 7 outputs the signal subjected to the still image processing at the circled portion of f and the signal subjected to the moving image processing at the other shaded portions.

Next, consider a case where a motion vector signal is transmitted and motion correction is performed. As in Figure 3a, only the circled part is stationary,
Suppose that the others are moving to the left. At this time, since the read address of the motion compensation type frame memory 2 is shifted by Δt, its output is that when motion compensation is not performed d.
It is shifted to the left by Δt (FIG. 4g), and the portions other than the circles overlap with e. Therefore, if motion detection is performed by comparing e and g, it is determined that only the shaded portion of the circle as in h is a moving image, and the rest is a still image. That is, in the mix circuit 7, the signal subjected to the moving image processing is output in the circled portion of h, and the signal subjected to the still image processing is output in other cases.

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention However, in the above method, when motion compensation is performed, it is determined that a still portion is actually moving, moving image processing is performed, and image quality deteriorates. . This is because, in FIG. 4h, most of the screen moves in the same direction and the motion correction is performed even though the circled part is stationary.

The present invention solves the above problems and provides an apparatus for improving image quality by performing still image processing on a still portion even if motion correction processing is performed.

Means for Solving the Problems In order to achieve this object, the present invention uses, in addition to a motion compensation type frame memory, a frame memory unrelated to motion compensation in parallel, and A signal that does not apply is taken out at the same time, and a signal that does not apply motion compensation is also subjected to motion detection, so that a still image process is performed on a portion that is actually still even when motion compensation is performed.

Action Conventionally, when motion compensation is compensated, it is determined that all actually static portions are moving images, but in the present invention, a non-motion compensation frame memory is used in parallel with the motion compensation frame memory. By doing so, the still image processing can be performed on the still portion.

Embodiment An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows an image quality improving apparatus of the present invention. 1 is a non-motion correction type frame memory, 2 is a motion correction type frame memory, 3 is an input switching circuit, 4 is a motion detection circuit (I) 5.
Is a still image processing circuit, 6 is a moving image processing circuit, and 7 is a mixing circuit. The above is the same as the configuration of the conventional example of FIG.

Reference numeral 8 is a non-motion-correction type frame memory that does not perform motion correction even when it is necessary to perform motion correction. (II) and 10 are switches for switching between signals that have been subjected to motion correction and signals that have not.

If the switch 10 is set to the side a, the operation is exactly the same as that described in the conventional example, and the images of the respective parts are as shown in FIGS. In this case, if the motion correction is performed by the motion vector, the circled portion of the hatched portion in FIG.

When the switch 10 is set to the a side, an image in which motion compensation is not performed even if motion compensation by a motion vector is applied is output.
That is, it is judged by the motion detection (II) 9 that the part which is still still (circular part) as shown in (f) of FIG. At this time, the output of the motion compensation type frame memory 2 is as shown in FIG.
The still circled portion is determined to be a moving image. The output from the non-motion-correction type frame memory 8 can be taken out only by turning the switch to the side by the output signal of the motion detection circuit (II) 9 only for this circle portion. Therefore, when motion compensation is performed, the output of the motion compensation type frame memory 2 is selected for the portion moving in the same direction, and the output of the non-motion compensation type frame memory 8 is selected for the stationary portion. , If there is a still part even if motion compensation is applied, it is regarded as a still image,
The part that processes video will be reduced.

EFFECTS OF THE INVENTION As described above, according to the present invention, it is possible to perform still image processing by determining a still image if there is a still portion even if motion compensation is performed, which is extremely effective in improving image quality. There is a great one.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an image quality improving apparatus according to an embodiment of the present invention, and FIG. 2 is a block diagram showing the configuration of a conventional example.
FIG. 4 and FIG. 4 are front views of the screen for explaining the operation of FIG. 1 and FIG. 1 ... Non-motion correction frame memory, 2 ... Motion correction frame memory, 3 ... Input switching circuit, 4 ... Motion detection circuit (I), 5 ... Still image processing circuit, 6 ... Moving image processing circuit , 7 ... Mix circuit, 8 ... Non-motion compensation frame memory, 9 ... Motion detection circuit (II), 10 ... Switch.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yoshimichi Otsuka 1-10-11 Kinuta, Setagaya-ku, Tokyo Inside the Japan Broadcasting Corporation Broadcasting Technology Laboratory (72) Inventor Izumi ▲ Yoshi ▼ Nori Kinuta, Setagaya-ku, Tokyo 10-11, Japan Broadcasting Corporation Broadcasting Technology Laboratory

Claims (1)

[Claims] 1. An output of a first frame memory for storing one frame of a television signal is added to a second frame memory and a third frame memory, respectively, and the output of the second frame memory or the third frame memory and the received input. A television signal is switched by an input switching circuit to perform interframe interpolation processing to be an input signal of the first frame memory, and the second frame memory is a motion vector signal representing a motion amount between frames or fields of the image signal. In addition to configuring so that the image can be shifted according to
The output of the third frame memory is compared with the output signal of the input switching circuit subjected to interframe interpolation processing to detect a still portion of the image, and in the interframe interpolation processing,
The output of the third frame memory and the input television signal are switched in the stationary portion, and the output of the second frame memory and the input television signal are switched in the non-stationary portion. The still image processing is performed from the output signal of the input switching circuit by the signal which detects the moving part by comparing the signal to be fed back to the input switching circuit from the output of the frame memory with the output signal of the input switching circuit. The signal obtained by selecting or synthesizing the selected signal and the signal subjected to the moving image processing is obtained.
An image quality improving device characterized in that a still image portion can be processed even when there is a motion vector signal.