JP2000350095A - Video processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a video processor by which the display range of subtitles is detected and masking is executed only there by generating a mask image for covering blocks based on the density of specified color dots, which is decided to be the subtitles blocks, and compositing the mask image with the inputted image. SOLUTION: A signal fetching part 1 detects the 100% white dots, which seem to be subtitles telops and inputs the density information to a subtitles deciding part 2. The part 2 decides the block, where the subtitles are displayed based on density information inputted from the fetching part 1 and the timewise change situation of the density information and indicates a signal processing part 3 to generate a video signal for masking the area, where the subtitles blocks are distributed. The part 3 generates the mask image only for the subtitles areas, composites it with the video signal to be inputted and then outputs it to a monitor. Thus, the video where the subtitles telops are hidden by the mask image is displayed on the monitor.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video processing apparatus for masking a subtitle telop from a video signal such as a foreign image to make it invisible.

[0002]

SUMMARY OF THE INVENTION Video images of foreign films, etc., which are not dubbed but sound in their original language,
Generally, a caption telop is displayed at the bottom of the screen. However, for those who are trying to learn languages from foreign films, there is a drawback that if subtitles are present, their consciousness will be drawn to the subtitles and they will not be able to concentrate on language conversation.

[0003] A device for masking subtitles is conceivable. However, since the range in which subtitles are displayed by video is not constant, it is not possible to determine an appropriate masking range in advance. When a wide range of mask images is generated, there is a disadvantage that the range becomes too wide to enjoy the movie itself.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a video processing device capable of detecting a subtitle display range and masking only the display range.

[0005]

SUMMARY OF THE INVENTION The present invention detects whether each dot of an image input as a video signal is a specific color which is a display color of a subtitle, and divides the image into a plurality of blocks. Specific color density detecting means for detecting the density of the specific color dot for each divided block; and displaying subtitles based on the density of the specific color dot detected a plurality of times at predetermined intervals by the specific color density detecting means. A subtitle block determining unit that determines a subtitle block that is a block, and a synthesizing unit that generates a mask video covering the block determined to be the subtitle block and synthesizes the mask video with the input video. Features.

The display color of the caption is generally 100% white, but any other color can be detected by setting it in advance. The video is divided into blocks, and the density of a specific color, which is the display color of the caption, is detected. The blocks may be divided into rectangles of s × m dots or linearly of 1 × m. The linear division is performed because a video signal is input as a scanning line signal, so that real-time processing can be performed without using a memory. In addition, by making 1 × 1 dots into one block, dot-by-dot processing is also possible. This is also included in the present invention.

[0007] Since the video signal is a moving image,
The color (including luminance) of each dot and each block changes smoothly with time. However, since the subtitle is a still image that is displayed and deleted as needed, the color does not change during display, and the color changes discontinuously at the display and deletion timing.
Based on these characteristics, it is determined which block is the subtitle block displaying the subtitle, and a mask image for masking the subtitle block is generated and combined with the original image. This makes it possible to generate a mask image of an appropriate size and mask the subtitle even if the display area of the subtitle is not known in advance.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A video processing apparatus according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of the video processing device. This video processing device detects captions displayed in white (100% white) and covers the display area with a gray mask image.
In a normal video image, the color (including brightness) constantly fluctuates, and the change is continuous. On the other hand, the subtitle telop is a still image that is displayed for a fixed period of time. And the color changes discontinuously during erasure. Also, it is rare for ordinary video images that many dots of exactly the same color (100% white) are distributed. This video processing device determines a display area of a caption telop using these characteristics of a caption telop, and displays a mask image only in this area.

This apparatus has a signal acquisition unit 1, a subtitle determination unit 2,
And a signal processing unit 3. The video signal is input to the signal capturing unit 1 and the signal processing unit 3. The signal capturing unit 1 detects a 100% white dot that is considered to be a caption telop, and inputs the density information to the caption determining unit 2. The density information is obtained by dividing the entire screen into blocks each including one to a plurality of dots, calculating the density occupied by 100% white dots for each block, and calculating the density of the blocks in which 100% white dots are distributed at a certain density or higher. This is information in which “1 = candidate block” is assigned and “0” is assigned to other blocks. In this embodiment, one block is composed of one dot in the vertical direction (one scanning line) × m dots in the horizontal direction (1 / n of the total number of dots in one scanning line).

[0010] The subtitle determination unit 2 determines the block in which the subtitle is displayed based on the density information input from the signal acquisition unit 1 and the temporal change of the density information.
It instructs the signal processing unit 3 to generate a video signal that masks the area where the subtitle blocks are distributed. The signal processing unit 3 generates a mask image of only the above-mentioned subtitle area, combines this with a video signal to be input, and outputs it to a monitor.
As a result, an image in which the caption telop is hidden by the mask image is displayed on the monitor.

FIG. 2 is a block diagram of the signal acquisition unit 1.
FIG. 3 and FIG. 4 are diagrams showing signals of each unit of the signal acquisition unit 1. In FIG. 2, the signal capturing unit 1 includes a video signal Vid
eo, which is input to the comparator 10, the H-pulse generator 1
7, input to the n-pulse generator 18 and the V-pulse generator 19. The comparison input terminal of the comparator 10 has a threshold voltage T
h1 is supplied, the output terminal outputs "H" when the input video signal is higher than the threshold voltage Th1, and outputs when the input video signal is lower than the threshold voltage Th1. "L" is output from the terminal. FIG.
As shown in (A) and (B), the threshold voltage Th1 is 1
The level is set to a level close to 00% white, and "H" is output only for subtitles and some pure white video portions. In a general video, there are not many such 100% white dots, and therefore, the subtitle portion has a higher “H” density than other portions.

The output of the comparator 10 opens and closes a switch 11. That is, the switch 11 from which the comparator 10 outputs “H” is turned on, and the switch 11 is off while the comparator 10 is outputting “L”. The switch 11 is a semiconductor switch composed of an FET or the like, and has a response speed of about 0.1 μS. This makes it possible to obtain a horizontal resolution of about 400 dots in a video signal that is a continuous time signal. The output of the comparator 10 becomes “H” and the switch 11
Turns on, current flows from the constant current source 9 into the averaging circuit C. The averaging circuit C is composed of a capacitor,
Charges corresponding to the time when the switch 11 is turned on are accumulated, and a voltage between terminals is generated according to the charge (see FIG. 4). This voltage is supplied to the comparator 14.

On the other hand, the H-pulse generator 17 converts the input video signal Video into a horizontal synchronization signal H.H. SYNC is detected, and a pulse (H pulse) synchronized with the SYNC is output (see FIG. 4). The n-pulse generation unit 18 is configured to
7 from the input video signal Video to the horizontal synchronization signal H.H. SYNC is detected, and the horizontal synchronizing signal H.SYNC is detected. SYN
Pulses (n pulses) are output at intervals obtained by equally dividing the time interval of C (1/1575 seconds) by n (see FIG. 4). That is, the n pulses are pulses that divide the time of one scanning line into blocks. Further, the V pulse generating unit 19 outputs the vertical synchronizing signal V.V. SYNC is detected and a pulse (V pulse) synchronized with this is output.

The H pulse is supplied as a latch pulse to a shift register 15 described later, and is input to the CPU 21 as an interrupt signal. The V pulse is used for a PIO to be described later.
16 is input as a field delimiter signal. Also,
The n pulse is supplied to the shift register 15 as a shift pulse (see FIG. 4), and is output to a reset switch 13 connected in parallel to the averaging circuit C via a delay 20 (see FIG. 4). ). The delay 20 is a delay having a pulse width of about n pulses as shown in FIG. 4, and is used to delay the reset of the averaging circuit C until the shift register 15 samples the output of the comparator 14 using the n pulses as a trigger. belongs to.

The averaging circuit C accumulates charges corresponding to the "H" time for each block, and outputs a voltage corresponding to the charges (see FIG. 4). When the time of the block has elapsed and the next block is reached, the averaging circuit C is reset by closing the reset switch 13. The voltage of the averaging circuit C is output to the comparator 14. The comparison input terminal of the comparator 14 is supplied with a threshold voltage Th2 for determining a candidate block having a high possibility of a block displaying a caption based on the density of the white dot, and averaging the threshold voltage Th2. When the input voltage of the conversion circuit C becomes higher than this, the comparator 14 outputs an "H" signal (see FIG. 4).

The shift register 15 has a shift pulse (n
When a pulse is input, the output of the comparator 14 is sampled and the contents of the register are shifted. Thereafter, when a reset pulse (n pulses delayed by the delay 20) is input to the reset switch 13, the charge of the averaging circuit C is discharged, and the output voltage and the output of the comparator 14 are reset.

The shift register 15 is an n-stage shift register, and includes a comparator 10 → a switch 11 → an averaging circuit C →
The sequence from the comparator 14 to the shift register 15 is the video signal V
The above operation is repeated n times during one scan line of the video. Then, after n operations, the H pulse generation unit 1
When an H pulse is input from 7, the value of the n-stage shift register is transferred to the PIO 16 as an n-bit signal.
At the same time, the CPU 21 outputs the n-bit signal latched in the PIO 16 in response to the interruption by the H pulse to the RAM 23.
Take in.

The V pulse output from the V pulse generator 19 is also input to the CPU 21 via the PIO 16. C
When the V pulse is input, the PU 21 changes the current video signal to the V.PU. It is possible to know that the position is a SYNC position, which enables processing in units of fields and frames.

In FIG. 2, the CPU-BUS to which the PIO 16 is connected includes a CPU 21 as a microcomputer unit,
In addition to the RAM 23, a ROM 22 and a hard disk 24 are connected. The ROM 22 stores a start-up program for starting the apparatus. A subtitle determination program is stored in the hard disk 24, and is read into the RAM 23 at startup. This microcomputer section corresponds to the caption determination section 2 in FIG. It is to be noted that the hard disk 24 is unnecessary if the program is stored in a ROM.

The subtitle determination operation of the CPU 21 will be described with reference to the memory configuration diagram of FIG. 5 and the flowchart of FIG. In FIG. 5A, n × 48 is stored in the RAM 23.
A 3-bit frame memory is set for 20 frames. Each frame memory is a storage area for storing one block of density information of n blocks for one scanning line. N
In the TSC signal, the number of scanning lines in one frame is 525, but 483 is sufficient in the vertical direction because about 40 of them are in the blanking period. Each frame memory is identified by a frame pointer i.

In FIG. 6, when this apparatus is started, first, an initialization operation is executed (s1). The initialization operation is a process of clearing the RAM 23, reading the program shown in the figure, and setting 1 to the frame pointer i. After that, it waits at s2 until the video signal Video is input. The input of the video signal can be detected by interruption by the H pulse.

When a video signal is input, density information for one frame is fetched (s3). The start of the frame
The determination can be made by inputting a V pulse from the PIO 16 or the like. 1 every time the H pulse is input
The density information of the scanning lines is fetched, and the density information of all the scanning lines except the blanking period is written in the frame memory, and then the process of s3 is ended.

The processing of s3 is executed for 20 frames (s
4), instead of capturing from consecutive frames,
At an interval of about 1/2 second (15 frames) (s6),
take in. After the capture, it is determined which block is a subtitle block for displaying subtitles (s7). This determination is performed as follows. Since the video signal is a moving image, the color (including brightness) of each block (dot) continuously changes with time. On the other hand, since the caption is displayed in a stationary state, the caption is similarly 100% white even after a lapse of time. When the subtitle is deleted, the color of the block changes discontinuously. This discontinuous change is only in the subtitle part, unlike the case where the entire screen is switched. Also, since subtitles are generally displayed in one horizontal line or one vertical line, blocks of a specific color (in this case, 100% white) are concentrated in a certain section. Based on such characteristics, a time-series frame memory is inspected to determine a subtitle block (see FIG. 5B). When the subtitle block is determined, as shown in FIG. 5B, the smallest rectangular area that covers all the subtitle blocks is determined and set as a mask area. However, as shown in the figure, a subtitle block that is determined as a subtitle block but is isolated from other subtitle blocks is ignored because it is a video portion that happened to be 100% white.

In FIG. 6, this mask area is stored in RAM
23 (s8), and outputs a signal specifying a mask area to the signal processing unit 3 to control the processing of the video signal (s9). This control is continued while the video signal is being input.

FIG. 7 is a block diagram of the signal processing unit. FIG. 8 is a diagram illustrating an example of an image of a video signal processed by the signal processing unit.

The video signal Video is input to one of the input terminals of the synchronizing signal separating circuit 31 and the video switch 33. An output terminal of the video switch 33 is connected to the monitor, and a mask image from the NTSC encoder 32 is input to the other input terminal.

The CPU 21 instructs the video chip 30 on the mask area and the color of the mask image. The video chip 30 receives the H.264 signal from the synchronization signal separation circuit 31. SYNC, vertical synchronization signal V. SYNC and subcarrier frequency f _sc are input. The video chip 30
Based on the synchronization signal, a video signal of the designated mask area is generated in synchronization with the video signal Video. This signal is an RGB signal of the designated color,
After being converted into an NTSC composite signal by the encoder 32, it is input to the video switch 33 (see FIG. 8B).

The video chip 30 also has a transparent signal YS
Is output. This transparent signal YS is a signal output when the video chip 30 itself does not create a video signal, and is output in an area other than the mask area. This YS signal is input to the video switch 33 as a switching signal. When the YS signal is input, the video signal Vi
Deo is output to the output terminal, and when the YS signal is not input, the mask image of the NTSC encoder 32 is output to the output terminal. As a result, the video output to the monitor becomes a video as shown in FIG. 8C in which the subtitle portion is masked.

In this embodiment, the signal capturing section is constituted by an analog circuit. However, the signal capturing section may be constituted by a digital circuit by digitizing a video signal.

In this embodiment, the blocks are 1 ×
Although m dots are used, a block may be formed over a plurality of scanning lines. Further, it is also possible to reconstruct a block extending over a plurality of scanning lines based on the contents stored in the frame memory.

Although the data input from the shift register 15 is 1-bit data, it may be multi-bit data according to the density of white dots.

If the size of the block is set to 1 × 1, it is determined in dot units, which is also included in the present invention.

Further, in this embodiment, the subtitle telop is displayed without white, but the telop may be of any color, and the object to be detected may be matched to that color.

By designating an area such as the lower third of the screen as an area where a caption telop should appear, a candidate area that appears in a completely different area can be excluded. This can facilitate the determination and eliminate erroneous determination.

In the above embodiment, the video signal input to the signal capturing unit 1 and the video signal input to the signal processing unit 3 have the same timing, and the subtitle determination unit 2 determines the subtitle area. For a few seconds to about 10 seconds, the video with the subtitle displayed is output to the monitor. Therefore, by providing a delay device for delaying the video signal by several seconds to about 10 seconds in the preceding stage of the signal processing unit 3,
Captions can be displayed from the beginning of the video signal.

It is not necessary that the delay time of the delay device and the determination time of the subtitle determination unit 2 completely match.

The subtitle determination unit may stop the operation completely after detecting the subtitle area. The subtitle determination unit always determines the presence or absence of the subtitle. May be canceled.

[0038]

As described above, according to the present invention, even if the display area of the caption is not known in advance, it is possible to detect the caption and synthesize the mask image to display the input image. You can enjoy videos with subtitles only in the original language.

[Brief description of the drawings]

FIG. 1 is a block diagram of a video processing apparatus according to an embodiment of the present invention;

FIG. 2 is a diagram showing a configuration of a signal acquisition unit of the video processing device.

FIG. 3 is a diagram for explaining the operation of a comparator of the signal capturing unit;

FIG. 4 is a diagram showing a signal of each unit of the signal acquisition unit;

FIG. 5 is a diagram showing a memory configuration of the subtitle determination unit.

FIG. 6 is a flowchart showing the operation of a caption determination unit of the video processing device.

FIG. 7 is a block diagram of a signal processing unit of the video processing device.

FIG. 8 is a diagram showing an example of a video signal processed by the signal processing unit.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Signal acquisition part, 2 ... Caption judgment part, 3 ... Signal processing part, 4
... delay circuit, 10 ... comparator, 11 ... switch, C ... averaging circuit, 12 ... constant current source, 13 ... reset switch, 1
4 ... Comparator, 15 ... (n-stage) shift register, 16 ... P
IO, 17 ... H pulse generator, 18 ... n pulse generator,
19 V pulse generator, 20 delay, 21 CP
U, 22 ROM, 23 RAM, 24 hard disk, 30 video chip, 31 synchronization signal separation circuit, 3
2: NTSC encoder, 33: Video switch

Claims (1)

[Claims] 1. A method for detecting whether each dot of an image input as a video signal is a specific color which is a display color of a caption, dividing the image into a plurality of blocks, and performing the identification for each of the divided blocks. Specific color density detecting means for detecting the density of color dots; and determining a subtitle block which is a block for displaying subtitles based on the density of the specific color dots detected a plurality of times at predetermined intervals by the specific color density detecting means. A video processing apparatus comprising: a subtitle block determining unit; and a synthesizing unit that generates a mask video covering a block determined to be a subtitle block, and synthesizes the mask video with the input video.