JP2005229231A - Motion vector detecting apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a motion vector detecting apparatus which has simple constitution and has reduced misdetection of a motion vector more than before when the motion vector detecting apparatus is applied in particular to a video signal adopting the interlace system. <P>SOLUTION: The motion vector detecting apparatus detects a motion vector V by switching a different detection reference for the motion vector V between a preceding field and a preceding frame on the basis of an inter-field difference DFD1 and an inter-frame difference DFD2. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a motion vector detection device, and in particular, can be applied to a case where a motion vector is detected from an interlace video signal. The present invention switches the motion vector detection reference between the previous field and the previous frame based on the comparison result between the inter-field difference value and the inter-frame difference value, and detects the motion vector by a simple configuration. Thus, it is possible to further reduce erroneous detection of motion vectors.

  Conventionally, in high-efficiency encoding processing of video signals, motion vectors are used to improve the efficiency of inter-frame encoding processing. In television signal format conversion, motion using motion vectors is used. By the correction, the jerky movement due to the conversion of the number of fields is corrected to a smooth movement.

  In such motion vector detection, a video signal is blocked into motion vector detection units (that is, macro blocks) of m pixels × n lines (m and n are integers), and a motion vector is detected for each block. For example, there are a pattern matching method disclosed in Japanese Patent Laid-Open Nos. 55-162683 and 55-162684, a gradient method disclosed in Japanese Patent Laid-Open No. 60-158786, a phase correlation method, and the like. It is made to be used.

  When a motion vector is detected from an interlaced video signal using such various motion vector detection methods, conventionally, a motion vector is detected between fields to prevent deterioration in detection accuracy when the amount of motion is large. Has been made. That is, the amount of motion between frames is twice that of between fields, so that the magnitude of the motion vector detected between frames is twice that of the motion vector detected between fields. When the amount of motion is large, the motion vector detection accuracy deteriorates.

  However, in the case of an interlaced video signal, the center of gravity of the video is shifted by ½ line in the vertical direction between the even field and the odd field, so that when the motion vector is detected between the fields, the center of gravity needs to be corrected in advance. However, when the center of gravity is corrected, the detection accuracy is deteriorated when the amount of motion is small. Specifically, even when the motion is “0” as in a still image, for example, the inter-field difference value cannot be the value “0” due to the centroid correction, which causes a detection error. Particularly in motion vector detection using the gradient method, by calculating a motion vector by calculation, the inter-field difference value does not become “0” even for a still image, and when there is an edge component of the image, A motion vector other than the value “0” is also detected in the still image.

As one method for solving this problem, a method of detecting a motion vector by converting an interlaced video signal into a progressive video signal can be considered. However, in the conversion process to the progressive method, the interpolation process is switched between the moving image and the still image, and the video signal of the previous field is used for the still image, but the moving image is the same as the center of gravity correction after all. Thus, even when the motion vector is detected after being converted into a progressive video signal, the motion vector is erroneously detected when the motion amount is small.
Japanese Unexamined Patent Publication No. 55-162683 Japanese Patent Laid-Open No. 55-162684 JP 60-158786 A

  The present invention has been made in consideration of the above points, and when detecting a motion vector with an interlaced video signal, it is possible to further reduce detection of a motion vector by a simple configuration as compared with the conventional case. It is intended to propose a motion vector detection device that can be used.

  In order to solve such a problem, the invention of claim 1 is applied to a motion vector detecting device for detecting a motion vector of an interlaced video signal, and outputs the video signal of the previous field by delaying the video signal by one field. 1 field delay means, 1 frame delay means for delaying the video signal by one frame and outputting the video signal of the previous frame, and detecting the inter-field difference value of the video signal based on the video signal and the video signal of the previous field A difference calculation means between fields, a difference calculation means between frames for detecting an inter-frame difference value of the video signal based on the video signal and the video signal of the previous frame, and an inter-field difference value and an inter-frame difference value. Comparing means that compares and outputs the comparison result, and when a predetermined condition is satisfied based on the comparison result Signal switching means for outputting the video signal of the previous frame instead of the video signal of the previous field, and movement between fields or between frames based on the video signal of the previous field or the video signal of the previous frame output from the signal switching means Motion vector detecting means for detecting a vector, and satisfying a predetermined condition is a case where the inter-frame difference value is smaller than the inter-field difference value.

  According to a second aspect of the present invention, in the configuration of the first aspect, centroid correction means for correcting a shift of the centroid of the video between adjacent fields of the video signal and outputting the same to the one-field delay means is provided.

  According to the configuration of the first aspect, the one-field delay means is applied to the motion vector detecting device for detecting the motion vector of the interlaced video signal and outputs the video signal of the previous field by delaying the video signal by one field. 1 frame delay means for delaying the video signal by 1 frame and outputting the video signal of the previous frame, and a field difference for detecting the inter-field difference value of the video signal based on the video signal and the video signal of the previous field Computation means, inter-frame difference computation means for detecting the inter-frame difference value of the video signal based on the video signal and the previous frame video signal, and the comparison result between the inter-field difference value and the inter-frame difference value Based on the comparison result and the video signal of the previous field when a predetermined condition is satisfied. Instead of the signal switching means for outputting the video signal of the previous frame, and the motion for detecting the motion vector between fields or between frames based on the video signal of the previous field or the video signal of the previous frame output from the signal switching means A vector detection means, and satisfying a predetermined condition is a case where the inter-frame difference value is smaller than the inter-field difference value. Based on this, it is possible to detect the motion vector by switching the motion vector detection criterion between the previous field and the previous frame, and it is possible to further reduce the erroneous detection of the motion vector with a simple configuration as compared with the conventional case.

  According to a second aspect of the present invention, in the first aspect, the center of gravity correction means for correcting the shift of the center of gravity of the image between adjacent fields of the video signal and outputting the same to the one-field delay means is provided. Accordingly, it is possible to prevent erroneous detection of a motion vector due to a shift in the center of gravity of the video between fields.

  According to the present invention, when a motion vector is detected with an interlaced video signal, it is possible to further reduce erroneous detection of the motion vector with a simple configuration as compared with the conventional art.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.

(1) Configuration of Embodiment FIG. 1 is a block diagram showing a motion vector detection apparatus according to Embodiment 1 of the present invention. The motion vector detection device 1 is applied to, for example, a format conversion device, a coding device based on MPEG (Moving Picture Experts Group), a noise reduction device, and the like, and uses a motion vector V from a luminance signal Y constituting an interlaced video signal. Is detected and output. In the detection of the motion vector, not only when the luminance signal Y is used, but the color difference signals RY and BY can also be used.

  In this motion vector detection device 1, the pre-filter 2 is a two-dimensional low-pass filter, which receives the luminance signal Y and suppresses high frequency components, thereby preventing erroneous detection of motion vectors due to noise. The pre-filter 2 performs so-called center-of-gravity correction processing by switching the tap coefficient in the vertical direction between the even field and the odd field. The pre-filter 2 is not limited to a low-pass filter, and a band-pass filter can be applied.

  The 1-field delay circuit 3 receives the luminance signal Y from the pre-filter 2 and outputs it after being delayed by a period of 1 field.

  On the other hand, the pre-filter 4 is a two-dimensional low-pass filter, which receives the luminance signal Y and suppresses high frequency components, thereby preventing erroneous detection of motion vectors due to noise. The pre-filter 4 is not limited to a low-pass filter, and a band-pass filter can also be applied.

  The 1-frame delay circuit 5 receives the luminance signal Y from the pre-filter 4 and outputs the delayed signal by a period of 1 frame.

  The difference comparison circuit 6 is based on the luminance signal Y1 output from the one-field delay circuit 3, the luminance signal Y2 output from the one-frame delay circuit 5, and the luminance signal Y of the current field. Calculate and compare values.

  That is, FIG. 2 is a block diagram showing the difference comparison circuit 6 and its peripheral circuits. The difference comparison circuit 6 inputs the luminance signal Y of the current field and the luminance signal Y1 of the previous field to the difference calculation circuit 11, and The luminance signal Y of the current field and the luminance signal Y2 of the previous frame are input to the difference calculation circuit 14.

  The difference calculation circuit 11 calculates and outputs a difference value of pixel values between corresponding pixels between the luminance signal Y of the current field and the luminance signal Y1 of the previous field. The absolute value conversion circuit 12 calculates and outputs the absolute value of the difference value output from the difference calculation circuit 11. The cumulative calculation circuit 13 calculates a sum of absolute differences obtained by adding the absolute values output from the absolute value conversion circuit 12 for each macroblock, and outputs this as the inter-field difference value DFD1.

  On the other hand, the difference calculation circuit 14 calculates and outputs a difference value of pixel values between corresponding pixels between the luminance signal Y of the current field and the luminance signal Y2 of the previous frame. The absolute value conversion circuit 15 calculates and outputs the absolute value of the difference value output from the difference calculation circuit 14. The cumulative calculation circuit 16 calculates the sum of absolute differences obtained by adding the absolute values output from the absolute value conversion circuit 15 for each macroblock, and outputs this as the inter-frame difference value DFD2.

  Note that the sum of squares of the difference values may be applied to the inter-field difference value DFD1 and the inter-frame difference value DFD2.

  The comparison circuit 17 compares the inter-field difference value DFD1 output from the cumulative calculation circuit 13 with the inter-frame difference value DFD2 output from the cumulative calculation circuit 16, and outputs the comparison result to the signal switching circuit 7.

  The signal switching circuit 7 outputs the luminance signal Y1 of the previous field input from the one-field delay circuit 3, and if the predetermined condition is satisfied based on the comparison result output from the comparison circuit 17, this luminance The output of the signal is switched, and the luminance signal Y2 of the previous frame is output instead of the luminance signal Y1 of the previous field. That is, when the interframe difference value DFD2 is smaller than the interfield difference value DFD1, based on the comparison result output from the comparison circuit 17, the signal switching circuit 7 replaces the luminance signal Y1 of the previous field with the previous frame. A luminance signal Y2 is output. When outputting the luminance signal Y2 of the previous frame, a determination signal S1 for identifying the luminance signal Y2 of the previous frame is also output.

  The motion vector detection circuit 8 sequentially detects the motion vector V of the luminance signal Y of the current field based on the luminance signal Y1 of the previous field output from the signal switching circuit 7 or the luminance signal Y2 of the previous frame by the gradient method. And output. In this embodiment, macroblocks are sequentially set in the luminance signal Y of the current field, and the motion vector V is detected in units of macroblocks. This macroblock is set to a size of 8 pixels × 8 lines. It is made so that. Accordingly, in this embodiment, the motion vector is detected by switching the motion vector detection criterion between the previous field and the previous frame based on the comparison result between the inter-field difference value DFD1 and the inter-frame difference value DFD2. .

  If the motion vector detected by the motion vector detection circuit 8 is a motion vector detected between frames based on the determination signal S1 output from the signal switching circuit 7, the motion vector correction circuit 9 By halving the vector size, a motion vector between frames is converted into a motion vector V between fields, and the motion vector detected by the motion vector detection circuit 8 is converted between fields. If it is a motion vector detected in step (1), this motion vector is output as it is. The motion vector V output from the motion vector correction circuit 9 is applied to, for example, field interpolation.

  Accordingly, in this embodiment, the 1-field delay circuit 3 constitutes 1-field delay means for delaying the video signal by 1 field and outputs the video signal of the previous field, and the 1-frame delay circuit 5 converts the video signal to 1 1 frame delay means for outputting the video signal of the previous frame by delaying the frame is constituted, and the difference calculation circuit 11, the absolute value conversion circuit 12 and the cumulative calculation circuit 13 are based on the video signal and the video signal of the previous field. The inter-field difference calculating means for detecting the inter-field difference value DFD1 of the signal is configured. The difference calculating circuit 14, the absolute value converting circuit 15 and the total calculating circuit 16 are based on the video signal and the video signal of the previous frame. The inter-frame difference calculating means for detecting the inter-frame difference value DFD2 of the signal is configured, and the comparison circuit 1 Constitutes a comparison means for comparing the inter-field difference value DFD1 and the inter-frame difference value DFD2 and outputting the comparison result, and the signal switching circuit 7 is used when the predetermined condition is satisfied based on the comparison result. The signal switching means for outputting the video signal of the previous frame instead of the video signal of the field is configured, and the motion vector detection circuit 8 is based on the video signal of the previous field or the video signal of the previous frame output from the signal switching means. A motion vector detection means for detecting a motion vector between fields or frames is configured.

(2) Operation of the embodiment In the above configuration, in the motion vector detection device 1, the luminance signal Y of the interlaced video signal to be processed is frequency-converted by the prefilter 2 in the even field and the odd field. After the center of gravity is corrected, the signal is delayed by one field period by the one-field delay circuit 3 and output as the luminance signal Y1 of the previous field.

  For the luminance signal Y1 of the previous field, the difference comparison circuit 6 calculates a difference value between corresponding pixels of the luminance signal Y of the current field, and calculates an inter-field difference value DFD1 for each macroblock.

  In parallel with this, the luminance signal Y is corrected by the prefilter 4 and then delayed by one frame period by the one frame delay circuit 5 and is output as the luminance signal Y2 of the previous frame. .

  With respect to the luminance signal Y2 of the previous frame, the difference comparison circuit 6 calculates a difference value between corresponding pixels of the luminance signal Y of the current field, and calculates an interframe difference value DFD2 for each macroblock.

  In this motion vector detection device 1, the inter-field difference value DFD1 and the inter-frame difference value DFD2 are compared by the difference comparison circuit 6, and the luminance signal Y1 of the previous field or the luminance signal of the previous frame is compared with the signal switching circuit 7 by the comparison result. Y2 is output to the motion vector detection circuit 8. This also allows the motion vector detection circuit 8 to detect the motion vector by switching the detection criterion of the motion vector between the previous field and the previous frame based on the comparison result between the inter-field difference value DFD1 and the inter-frame difference value DFD2. The motion vector V detected by the motion vector detection circuit 8 is corrected by the motion vector correction circuit 9 and output in conjunction with the switching of the detection reference.

  Accordingly, in the motion vector detection device 1, when the inter-frame difference value DFD2 is smaller than the inter-field difference value DFD1, a motion vector is detected and output between frames.

  That is, the inter-field difference value DFD1 does not fall to the value “0” even when the motion is “0” because the center of gravity is shifted by ½ in the field, whereas the inter-frame difference value DFD2 In that case, the value falls to “0”. In this case, a motion vector can be detected with higher accuracy between frames than between fields. Such a situation occurs not only when the movement is “0” but also when the movement is small.

  On the other hand, when the motion is large, the motion vector can be detected with higher accuracy between the fields. In this case, the inter-field difference value DFD1 is naturally changed to the inter-frame difference value DFD2. The value is smaller than that.

  Thus, as in this embodiment, when the inter-frame difference value DFD2 has a smaller value, if the motion vector is detected between frames, the motion vector can be detected with higher accuracy than in the past. it can.

  By applying the motion vector detection device 1 according to this embodiment to a television format conversion device that performs motion correction using motion vectors, the motion vector detection accuracy can be improved and image distortion can be reduced. . Further, in the scan converter, the resolution of a moving image can be increased by performing inter-field line interpolation processing using a motion vector, so that the motion vector detection device 1 of this embodiment is applied to increase the motion vector detection accuracy. Thus, the conversion image quality can be improved.

  Even in the region where the correlation in the line direction is small, the inter-frame difference value DFD2 is smaller than the inter-field difference value DFD1, and in this case, instead of detecting the motion vector between fields, a motion vector is calculated between frames. By detecting, a motion vector can be detected with high accuracy.

  Further, when the detection of motion vectors is switched between fields and between frames based on the comparison result of the inter-field difference value DFD1 and the inter-frame difference value DFD2, a method of providing motion vector detection means between fields and between frames is also provided. In this embodiment, the motion vector detection means is switched between the previous field and the previous frame for one system of motion vector detection means, thereby detecting a motion vector with a simple configuration. Can do.

(3) Effects of the embodiment According to the above configuration, the motion vector is switched by switching the motion vector detection reference between the previous field and the previous frame based on the comparison result between the inter-field difference value DFD1 and the inter-frame difference value DFD2. By detecting, it is possible to further reduce the erroneous detection of motion vectors by a simple configuration as compared with the prior art.

  Further, by detecting the motion vector by correcting the shift of the center of gravity of the video between the fields, it is possible to prevent erroneous detection of the motion vector due to the shift of the center of gravity of the video between the fields.

  In the above-described embodiment, the case where the inter-field difference value DFD1 and the inter-frame difference value DFD2 are detected in units of macroblocks has been described. However, the present invention is not limited to this, and other various sizes of block units are used. You may make it detect.

  In the above-described embodiment, the case where the motion vector is detected in units of macro blocks of 8 pixels × 8 lines has been described. However, the present invention is not limited to this, and the detection is performed in units of macro blocks of various other sizes. You may do it.

  In the above-described embodiment, the case where the motion vector is detected by the gradient method has been described. However, the present invention is not limited to this, and may be detected by another method such as a block matching method or a phase correlation method. .

  The present invention relates to a motion vector detection device, and can be applied to, for example, a format conversion device.

It is a block diagram which shows the motion vector detection apparatus which concerns on the Example of this invention. It is a block diagram which shows the difference comparison circuit of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Motion vector detection apparatus, 2, 4 ... Pre filter, 3 ... 1 field delay circuit, 5 ... 1 frame delay circuit, 6 ... Difference comparison circuit, 7 ... Signal switching circuit, 8 ... Motion vector detection circuit, 9... Motion vector correction circuit, 11, 14... Difference calculation circuit, 12, 15... Absolute value conversion circuit, 13, 16.

Claims (2)

In a motion vector detection device for detecting a motion vector of an interlaced video signal,
1 field delay means for delaying the video signal by 1 field and outputting the video signal of the previous field;
1 frame delay means for delaying the video signal by 1 frame and outputting the video signal of the previous frame;
An inter-field difference calculating means for detecting an inter-field difference value of the video signal based on the video signal and the video signal of the previous field;
An inter-frame difference calculating means for detecting an inter-frame difference value of the video signal based on the video signal and the video signal of the previous frame;
A comparison means for comparing the inter-field difference value and the inter-frame difference value and outputting a comparison result;
Based on the comparison result, signal switching means for outputting the video signal of the previous frame instead of the video signal of the previous field when a predetermined condition is satisfied;
Motion vector detection means for detecting a motion vector between fields or between frames based on the video signal of the previous field or the video signal of the previous frame output from the signal switching means,
When the predetermined condition is satisfied,
The motion vector detection device, wherein the inter-frame difference value is smaller than the inter-field difference value. The motion vector detection device according to claim 1, further comprising: a center of gravity correction unit that corrects a shift of the center of gravity of an image between adjacent fields of the video signal and outputs the corrected image to the one-field delay unit.

Images