JP2004208117A - Image quality deterioration detection system when compositing caption - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that it is erroneously determined that image quality deterioration occurs when a caption such as a news flash is composited. <P>SOLUTION: A video signal (a) is inputted to a feature quantity data generating section 1, an area designating section 7 designates an area, and feature quantity data A representing an image feature is generated. The generated feature quantity data A is multiplexed with the video signal (a) by a data multiplexer section 2 and is sent as a video signal b to a broadcast facility 3. The video signal b is demultiplexed by a data demultiplexer section 8 into the video signal (a) and the feature quantity data A, and the video signal (a) is composited with caption data and then remultiplexed with the demultiplexed feature quantity data. In a monitoring part, the data demultiplexer section demultiplexes a video signal c into the feature quantity data A and the video signal c again, and the video signal c is inputted to a feature quantity data generating section 10. The feature quantity data generating section 10 generates second feature quantity data C from an image of the video signal c. The demultiplexed feature quantity data A and the generated feature quantity data C are inputted to a comparison section 5, and an alarm is outputted when a difference is detected between the feature quantity data A and the feature quantity data C. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a system for detecting image quality deterioration of broadcast equipment in a broadcasting station, and more particularly to a system for detecting image quality deterioration that occurs when subtitles are combined with video in a broadcast equipment.
[0002]
[Prior art]
A conventional monitoring system will be briefly described with reference to the drawings.
(Our patent application, Japanese Patent Application No. 2001-332445, "Image quality deterioration detection method")
[Input part]
The video signal a is input to the feature data generating unit 1 to generate feature data A representing the features of the image. The generated feature amount data A is multiplexed on the video signal a by the data multiplexing unit 2 and sent to the broadcasting equipment 3 as the video signal b.
[0003]
In the broadcasting equipment 3, the subtitle is synthesized with the video signal b and transmitted to the monitoring part as the video signal b '.
[Monitoring part]
The video signal b ′ that has passed through the monitoring section is input to the data separation unit 4, separated into feature data A and a video signal a ′, and the video signal a ′ is input to the feature data generation unit 10.
[0004]
The feature data generator 10 generates feature data B from the image of the video signal a ′. The feature data A and the generated feature amount data B are input to the comparing unit 5, and the comparing unit 5 compares the two feature amount data and outputs an alarm indicating occurrence of image quality deterioration when a difference is detected.
Separately, Japanese Patent Application Laid-Open No. 6-178223 discloses a technique for reliably detecting character components such as subtitles contained in a video portion without erroneous detection without fail. Such technologies exist. Hereinafter, this technique will be briefly described.
[0005]
The input video signal is input to the adder, and the video signal output from the memory is also input to the adder. The adder adds the luminance video signal, which is a luminance component of the video signal, to the luminance video signal output from the memory and delayed by one field, and outputs the result. The signal output from the adder is input to a 係数 coefficient unit and is halved. As a result, the luminance video signal is subjected to the average value processing. Further, the averaged luminance signal output from the 計数 counter is input to the memory and averaged repeatedly to obtain a luminance video signal in which the character component of the subtitle portion is emphasized. By comparing the luminance video signal in which the character component is emphasized with the level comparator, the subtitle portion can be extracted without apologetic.
[0006]
[Problems to be solved by the invention]
In the related art, when subtitles such as breaking news are combined, there is a problem that it is determined that image quality has deteriorated. That is, the feature amount data B generated from the video signal b 'has a part always different from the feature amount data A to be compared by the amount generated from the video signal b' obtained by combining subtitles with the video signal b. Therefore, strictly speaking, it is always determined that the image quality is degraded and an alarm is generated. This can be dealt with by making the width of the comparison judgment loose so as not to judge the subtitles as image quality deterioration. However, this is one factor that hinders the improvement of the accuracy of the image quality deterioration judgment.
[0007]
In particular, in recent years, in order to display character information that cannot be conveyed by conventional subtitle synthesis, such as typhoon information and traffic information, the image has been reduced and a large amount of L-shaped portions formed at the periphery of the screen, such as electric light displays, have been displayed. Increasingly, character information is displayed or displayed while moving the character information vertically and horizontally. In such a case, with the conventional image quality deterioration detection method, even though the image itself has not deteriorated, the overall screen is significantly changed, so an image quality deterioration alarm is generated every time the subtitle is displayed. This would cause a malfunction.
[0008]
SUMMARY OF THE INVENTION An object of the present invention is to provide a system for correctly detecting image quality degradation at a monitoring point even when subtitles such as breaking news are synthesized by broadcasting equipment in a monitoring section for image quality degradation in order to solve the above-described problem. It is in.
[0009]
[Patent Document 1]
JP-A-6-178223
[Means for Solving the Problems]
In view of the above problems, the image quality degradation detection system of the present invention
A first feature data generating unit for generating first feature data from the video signal; a data multiplexing unit for multiplexing the first feature data on a blanking signal of the video signal; and the first feature data. A subtitle synthesizing unit that performs subtitle synthesis on the multiplexed video signal, a data separation unit that separates the first feature amount data from the subtitle synthesized video signal, and a second An image quality degradation detection system comprising: a second feature amount data generating unit that generates feature amount data; and a comparing unit that compares the first feature amount data and the second feature amount data to detect image quality degradation. The first feature value data generating means and the second feature value data generating means each include an area designating means for designating an area where the subtitle of the original video signal is not combined It is characterized.
[0011]
Further, as a second configuration example of the present invention, a blanking signal replacing unit is provided on the output side of the monitoring section, and the feature amount data of the video signal in which the feature amount data is multiplexed is used for broadcasting. It is characterized by being replaced with data such as broadcast.
[0012]
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment of an image quality deterioration detection system according to the present invention will be described in detail with reference to the drawings.
[0013]
FIG. 1 is a configuration diagram showing an embodiment of the present invention. The feature is that an area designating unit 7 for designating an area for generating feature quantity data is added to the feature quantity data generating unit 1 having a conventional configuration. A region where subtitles are not combined is selected by the region specifying unit 7, and this region is set as a monitoring target.
[0014]
By monitoring a portion that does not change even when subtitle synthesis is performed in this way, image deterioration detection can be performed correctly (FIG. 2).
[0015]
In the embodiment shown in FIG. 1, image quality deterioration occurring in a monitoring section is detected by generating feature amount data before and after a monitoring section and comparing the data. A subtitle synthesizing unit 6 is provided in the monitoring section, and synthesizes subtitles such as breaking news with the input video. Captions are synthesized only when necessary, and the control signal is used as a caption synthesis signal.
[0016]
Next, the configuration will be described separately for an input portion, a subtitle portion, and a monitoring portion.
[Input part]
When the video signal a is input to the feature data generation unit 1 and subtitle synthesis is performed at the timing specified by the subtitle synthesis signal, an area is specified by the area specifying unit 7 from the image of the video signal a, Is generated. At this time, the designated area is an area where subtitles are not combined.
[0017]
The generated feature amount data A is multiplexed on the video signal a by the data multiplexing unit 2 and sent to the broadcasting equipment 3 as the video signal b.
[Subtitle part]
Depending on the method of subtitle synthesis, the feature amount data A multiplexed on the video signal a may be damaged. Therefore, the subtitle synthesis is performed after the multiplexed feature amount data A is once separated according to the following procedure.
[0018]
The video signal b on which the feature data A is multiplexed is input to the data separation unit 8 in the subtitle part. The data separation unit 8 separates the video signal b into the video signal a and the feature data A, and sends the video signal a to the caption synthesis unit 6 and the feature data A to the data multiplexing unit 9 described later.
[0019]
The subtitle synthesizing unit 6 synthesizes subtitle data with the video signal a at the timing specified by the subtitle synthesizing signal, and sends out the video signal c to the data multiplexing unit 9. The data multiplexing unit 9 again multiplexes the feature amount data A separated by the data separation unit 8 onto the video signal c in which the subtitles are combined.
[0020]
Thus, the video signal c becomes a video signal in which the feature data A is multiplexed correctly even after the subtitle synthesis.
[Monitoring part]
The video signal c that has passed through the monitoring section is input to the data separation unit 4, separated into feature data A and the video signal c, and the video signal c is input to the feature data generation unit 10.
[0021]
The feature data generator 10 generates second feature data C from the image of the video signal c. At this time, the area of the image for generating the feature amount data C is specified by the area specifying unit 7 provided in the monitoring part. An area specified in the same manner as the input part is an area where subtitles are not combined. The separated feature data A and the generated feature amount data C are input to the comparing unit 5, and the comparing unit 5 compares the two feature amount data and outputs an alarm indicating occurrence of image quality deterioration when a difference is detected.
[0022]
Subsequently, the operation of the present invention will be described for each of the input part, the subtitle part, and the monitoring part with reference to the time charts shown in FIGS.
[Input part] (Fig. 3)
The caption synthesis signal indicates whether or not caption synthesis is performed. In this example, subtitles are synthesized from 12 frames to 14 frames of the video signal a. The video signal a is represented for each frame, and here, 10 to 19 frames are shown.
[0023]
The feature amount generation process 11 performs a process of generating feature amount data for each frame from the video signal a. It also shows that subtitles are synthesized in 12 to 14 frames by the subtitle synthesis signal 8. In these frames, the area in which the feature data A is generated is changed to a portion where subtitles are not combined. The feature amount multiplexing process 12 indicates a processing time for combining the feature amount data A with the video signal a. Within this time, the feature data A is multiplexed with the blanking signal of the video signal a. The video signal b is a video signal obtained by multiplexing the feature data A on the video signal a.
[0024]
In this example, 8 to 17 frames of feature data A are multiplexed on 10 to 19 frames of video signal a. This indicates that the multiplexed feature amount data A is delayed by two frames from the multiplexed video signal a.
[Subtitle part] (Fig. 4)
The caption synthesis signal 8 indicates the presence or absence of subtitle synthesis. In this example, subtitles are synthesized over 12 to 14 frames of the video signal b. The video signal b is represented for each frame. Here, 10 frames to 19 frames are shown. In each frame, eight to seventeen frames of feature data A are multiplexed, and among them, twelve to fourteen frames of feature data A are generated from an area where a subtitle is not combined.
[0025]
The feature amount separation process 13 indicates a processing time for separating the feature amount data A for each frame from the video signal b. At the end of the frame input of the video signal b, the feature amount data A is separated from the frame. The feature value multiplexing process 12 indicates a processing time for re-combining the feature value data A with the video signal a. Within this time, the feature data A is multiplexed with the blanking signal of the video signal a.
[0026]
The video signal c indicates a video signal output from the subtitle portion. In this example, 6 to 15 frames of feature data A are multiplexed on video signals c of 10 to 19 frames. This indicates that the multiplexed feature data A is delayed by 4 frames from the multiplexed video signal c.
[Monitoring part] (Fig. 5)
The caption synthesis signal 10 indicates whether or not caption synthesis is performed. In this example, subtitles are synthesized for 12 to 14 frames of the video signal c. The video signal c is represented for each frame, and here, 10 to 19 frames are shown. In each frame, feature amount data A of 6 to 15 frames are multiplexed. Subtitles are synthesized in 12 to 14 frames. In addition, the feature amount data A of the 12th to 14th frames is generated from a region where a subtitle is not combined.
[0027]
The feature amount generation processing unit 10 indicates a processing time for generating the feature amount data C for each frame from the video signal c. When the input of the frame of the video signal c is completed, the generation processing of the feature amount data C of the frame is performed. It also shows that subtitles are synthesized in 12 to 14 frames by the subtitle synthesis signal 10. In these frames, the area in which the feature amount data C is generated is changed to a portion where subtitles are not combined.
[0028]
The feature amount queue 14 stores the generated feature amount data C for a certain period. FIG. 6 shows a state in which the generated feature amount data C is stored and taken out after four frames.
[0029]
The feature amount separation processing 13 indicates a processing time for separating the feature amount data A for each frame from the video signal c. At the end of the input of the frame of the video signal c, a separation process of the feature data A from the frame is performed.
[0030]
Since the separated feature data A is delayed by four frames from the generated feature data C, the feature data C generated in the feature queue 14 is also delayed by four frames.
[0031]
The feature amount comparison processing 15 compares the feature amount data C delayed by four frames with the feature amount data A separated from the video signal. If the comparison results in a mismatch, an alarm indicating the occurrence of image quality degradation is output.
[0032]
Further, as another embodiment of the present invention, the entire image portion is reduced to a similar shape, and character information displayed while moving vertically and horizontally like an electronic display in an L-shaped frame formed on the periphery of the screen. There is a method of combining and using. In this case, in the monitoring part, feature amount data C ′ is generated from the video signal c ′ of the reduced screen to be monitored, and the feature amount data C ″ obtained by enlarging the feature amount data C ′ is compared with the feature amount data A. Is used to determine image quality degradation.
[0033]
【The invention's effect】
As described above, the image quality deterioration detection system according to the present invention when subtitles are synthesized has a malfunction in which, since subtitles and the like are superimposed on a part of an image, it is determined to be image quality deterioration when compared with the original image. It has the effect of preventing.
[0034]
Further, the image quality deterioration detection system of the present invention has an effect that, even when the entire image is reduced and subtitles are synthesized and displayed around the image, image deterioration can be correctly detected when compared with the original image.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of an image quality deterioration detection system according to an embodiment of the present invention.
FIG. 2 is an image diagram showing a comparison process between feature data A and feature data C in a comparison unit 5;
FIG. 3 is a time chart showing a series of operations in an input portion of the present invention.
FIG. 4 is a time chart showing a series of operations in a subtitle portion according to the present invention.
FIG. 5 is a time chart showing a series of operations in a monitoring part of the present invention.
FIG. 6 is an image diagram showing a comparison process between the feature data A and the feature data C ″ in the comparing unit 5 when an L-shaped frame is synthesized.
FIG. 7 is a block diagram illustrating a configuration example of a conventional image quality deterioration detection system.
FIG. 8 is an image diagram showing a feature amount data comparison process in a conventional example.
[Explanation of symbols]
REFERENCE SIGNS LIST 1 feature amount data generation unit 2 data multiplexing unit 3 broadcast equipment 4 data separation unit 5 comparison unit 6 subtitle synthesis unit 7 area designation unit 8 data separation unit 9 data multiplexing unit 10 subtitle synthesis signal 11 feature amount generation processing 12 feature amount multiplexing process 13 feature value separation process 14 feature value queue 15 feature value comparison process

Claims (6)

A first feature data generating unit for generating first feature data from the video signal; a data multiplexing unit for multiplexing the first feature data on a blanking signal of the video signal; and the first feature data. A subtitle synthesizing unit that performs subtitle synthesis on the multiplexed video signal, a data separation unit that separates the first feature amount data from the subtitle synthesized video signal, and a second An image quality degradation detection system comprising: a second feature amount data generating unit that generates feature amount data; and a comparing unit that compares the first feature amount data and the second feature amount data to detect image quality degradation. ,
Image quality degradation detection, characterized in that the first feature value data generation means and the second feature value data generation means each include an area designating means for designating an area where subtitles of the original video signal are not combined. system. A blanking signal replacing unit is provided on an output side of the monitoring section, and the feature amount data of the video signal in which the feature amount data is multiplexed is replaced with data such as a text broadcast used in broadcasting. 2. The image quality deterioration detection system according to 1. The feature amount data is generated by dividing an image into 64 (8 × 8), calculating an average value for each area, separating the signals into YCbCr signals, and performing an 8 × 8 two-dimensional DCT transform on each. The image quality degradation detection system according to claim 1 or 2, wherein The image quality deterioration detection system according to claim 1, wherein the monitoring section is a broadcast facility including a switcher or the like. The image quality deterioration detection system according to claim 1, wherein the monitoring interval is from output of a broadcast station to reception of a broadcast wave. The image quality deterioration detection system according to claim 1, 2 or 3, wherein the monitoring section extends from after the output of the broadcasting station to a relay point.

Images