JP2000134576A - Storage and reproduction device for title image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the quality of a restored title while decreasing the quantity of stored data on the title image for a title image storage/reproduction device which is used to an AV(audio video) mixer, etc., to edit the images. SOLUTION: This device includes a storage means which compresses and stores a title picture of a title image as picture data by the non-reversible encoding and also compresses and stores the title contour as pulling-out data by the reversible encoding, and a restoration means which restores the title contour from the pulling-out data which are compressed and stored by the storage means and synthesizes a restored title picture from the compressed picture data within the title contour to restore the title image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a title image storage / playback apparatus used for image editing in an AV (audio / video) mixer or the like.

[0002]

2. Description of the Related Art In editing a video, it is common to superimpose a title (character) at the start or end of the video, for example, on a real video. The equipment that performs the processing is called a titler.

[0003] In this titler, it is common to save frequently used titles in advance. However, if the title video data is saved as it is, the amount of saved data increases, and a large amount of data is stored in the memory. Is required. Therefore, the title video is encoded using an information compression method such as JPEG to reduce the amount of stored data. This JPEG encoding method can be expected to have a high information compression effect. When a video such as a landscape is restored, a certain level of quality can be obtained based on human visual characteristics. However, when a character video such as a title is restored, it can be obtained. However, because of the principle of the encoding method (encoding in block units), a so-called "whisker" is generated in the outline of the character, and the reproduction quality must be low.

On the other hand, if a title image is encoded using a run-length encoding method used for facsimile or the like, the restored image becomes good, but the run-length encoding method has poor compression efficiency. Therefore, for example, if a title video with a pattern is encoded as it is, the amount of stored data also increases. In this case, if the title image is saved as a so-called solid character title (one in which the outline of the character is colored), the amount of data to be saved can be reduced. Will not be able to create a title.

[0005] The present invention has been made in view of such a problem, and while reducing the amount of data to be stored for title images,
The purpose is to improve the quality of the restored title.

[0006]

FIG. 1 is an explanatory view of the principle according to the present invention. The title image storage / playback apparatus according to the present invention compresses and saves the title image of the title image as surface data by irreversible encoding, and compresses and compresses the title outline as lossy data as lossy data. A title storing means for restoring a title outline from the compressed extracted data saved by the saving means, and synthesizing the title image restored from the compressed plane data into the title outline to thereby convert the title video. Restoring means for restoring.

This storage / playback apparatus captures a title image by, for example, a photographing means, compresses and stores the title image of the title image as surface data by irreversible encoding, and stores the title outline of the title image. And compress the data by lossless encoding and store it. The restoring unit restores the title outline from the compressed extracted data stored by the storage unit, and combines the title outline restored from the compressed plane data into the title outline. The title outline restored based on the compressed and saved extracted data is clear without distortion because it is encoded by lossless encoding. For this reason, since the restored title surface has been encoded by irreversible encoding, even if a contour such as a “whisker” is generated in the contour, the distortion protrudes from the title contour. If this is done, it will be removed upon reproduction, and as a result, a high-quality title video can be restored.

The above-mentioned storage / playback apparatus can be configured to perform edge processing for adding a halftone edge along the title outline of the extracted data when saving the extracted data or restoring the extracted data. .
Thus, when restoring the title video, the video on the title surface and the background video are mixed at the halftone edge portion. In this way, the outline of the title video becomes familiar with the background video,
You can restore beautiful title videos.

The above-mentioned storage / playback apparatus is configured such that, when the surface data is stored or when the surface data is restored, a border attaching process for adding a border to the title surface of the surface data along the title contour is performed. be able to. As a result, the outline of the restored title video becomes clear.

In another aspect of the present invention, the above-mentioned plane data is obtained by compressing a pattern image to be attached to the title plane as plane data by irreversible encoding. The outline can be restored, and a pattern image restored from the compressed plane data can be combined with the title outline.

The storage / playback apparatus of the present invention compresses and saves the title image of the title image as surface data by irreversible encoding, and also compresses the title outline by extracting data as lossy data by lossless encoding. And restoring the title outline from the compressed extracted data, and restoring the title video by combining the title outline restored from the compressed plane data into the title outline. The present invention can also be realized by causing a computer to execute a computer-readable recording medium storing a program to be executed by a computer.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram schematically showing an overall configuration of an AV mixer equipped with a title video storage / playback apparatus as one embodiment of the present invention. In FIG.
U1 is a central processing unit that controls the entire apparatus, RAM2 is a random access memory that stores various data and provides, for example, a work memory area for title video processing, and ROM3 is a control program and various kinds of data. A read-only memory for storing tables and the like; a display 4 for an LCD (liquid crystal display) for displaying various setting data; a display driver 5 for driving and controlling the display 4; Reference numeral 6 denotes a hardware circuit and an operator group 8 for displaying a plurality of video signals by mixing or adding a title, and for mixing audio signals.
Is an operator group for various settings / selections arranged on the operation panel of the AV mixer, and an interface 7 is an interface for transmitting and receiving signals between the video / audio circuit 6 and the operator group 8 and the CPU 1 and the like. The title / still image file memory 9 is a memory that previously holds video data (compressed data) of various titles and still images as title image data.

FIG. 3 is an external view of the operation panel.
Various operation elements 8 and the display 4 are arranged on the operation panel. The operator group 8 is roughly divided into a video / audio switching switch button 20, a fade control switch button 21, a title / still image file selection switch button 22, and a manual video / sub video switching. 2 that is the control for
3, a switch button 24 for selecting a sub-picture, and a switch button 25 for various other operations used in processing such as title creation / display.

Here, a switch button 20 for switching between video and audio is provided with a normal button 201 for selecting a normal mode in which video and audio are separately switched according to the operation of a corresponding operating element, and a scene change. A color insert button 202 for selecting a color insert mode in which a color screen is inserted at a video switch at the time of video switching, and a voice linking button for selecting a voice linking mode for automatically switching audio in conjunction with video switching 203.

The switch buttons 21 for fade control include a white / black button 211, a source mode button 212, a total mode button 213, and an execute button 2.
Consists of fourteen. White / black button 211, source button 21
2. A total button 213 and an execution button 214 are buttons for selecting / executing a fade-out method when a title is superimposed on a video. That is, the source button 211 is a mode selection button for fading out only the video while leaving the title, the total button 213 is a mode selection button for fading out both the video and the title, and the white / black button 211 is a back screen at the time of fading out. The button for selecting (white / black) and the execution button are buttons for instructing execution of source fade or total fade.

The title / still image file switch button 22 is a button for designating a title / still image storage bank number (title number) stored in the title / still image file memory 9 for writing and reading. The group is composed of buttons A to C and 1 to 8, and a storage bank number is designated by a combination of these buttons. The memory card is a button for designating a memory card as a storage destination / source.

The fader 23 is an operation element for manually switching between the main image and the sub-image. The main image is moved as the operation knob is moved to the main image, and the sub-image is moved as the operation knob is moved to the sub-image. Is set at the position of the operation knob so that the mix ratio of the main video and the sub video is output so as to output more.

The auto take button 215 is used for the main video /
This is selected when switching the sub-pictures automatically. When this is turned on, the setting value manually operated by the fader 23 is ignored.

The sub-image switch button 24 is used as a sub-image for the video B input to the video / audio circuit 6.
And a title button 242 for selecting a title / still image.

Other buttons 25 include a cursor button 250 for moving a cursor, a value button 251 for setting a value, a setting button 252 for setting a mode to a setting mode, a menu button 253 for setting a mode to a menu mode, and Yes / Yes / No button 254 for selecting No, capture button 256, character extraction button 257, color processing button 258, transparency button 259, white button 261, black button 262, color button 263, merge button 264, PGM button 265, title in button 267, title out button 2
68, a WIPE button 269, a MIX button 270, and the like.

Here, the capture button 256 is a button for instructing to capture the video (the video of the title document) captured by the video camera 62 into the memory 35A in the video / audio circuit 6. Character extraction button 257
Is a button for instructing to binarize the captured video and extract a character portion.

The color processing button 258 is a button for instructing to perform color processing such as framing, shadowing, coloring, etc. on characters extracted from the title video. Transparency button 2
Reference numeral 59 denotes a button for setting transparency such as coloring when performing color processing.

White button 261, black button 262
Is a button for designating that the background of the title document from which characters are to be extracted is white or black, and a portion of the designated color is extracted from the video (color is removed). The title-in button 267 is a button for instructing insertion of a title into a video, and the title-out button 268 is a button for instructing deletion of a title from a video in which a title is inserted.

The merge button 264 is a button for instructing execution of a merge. The WIPE button 269 and the MIX button 270 are buttons for designating a video output method by wipe and mix, respectively.

FIG. 4 shows a detailed configuration of a video circuit portion related to video processing in the video / audio circuit 6 described above. 4, a video signal of a video A as a main video is input from a video deck 61 and a video signal of a video B as a sub video is input from a video camera 62 to the video circuit. In this embodiment, the main video A from the video deck 61 is a moving image or the like of a real photographed by the user, and the video B from the video camera 62 is a video of a document image on which a title (character) is written. .

The pre-processing unit 31 receives the input main image A
From the video signal, and sends the separated synchronization signal to the timing generator 33 and the remaining video signal to the input terminal of the modulator 41.

The pre-processing section 32 has the same function as the pre-processing section 31 and separates a synchronizing signal from a video signal of the input video B as a sub-video. The signal is input to the converter 34, digitized, and then input to the main controller 36.

When video B (actual video of a title document) is selected as a sub video by the video B button 241, the memory controller 36 converts the video data of video B input from the A / D converter 34 into one. The video data is read out immediately while being loaded into the frame memory 35A for each frame and output to the encoder 38 at the subsequent stage, and the built-in key circuit (shown in FIG. 5) creates key data based on the video data and generates a key. Output to the output unit 39. The video data sequentially taken into the frame memory 35A is the capture button 25
By pressing 6, only one frame can be captured, and at this time, the same captured video data (one frame) is repeatedly read.

On the other hand, when the title is selected as the sub-picture by the title button 242, the memory controller 36 stores the title / still image read out from the title / still image file memory 9 and restored in the frame memory 35A. The video data of the picture is transferred and stored, and the title data / still picture extraction data is transferred and stored in the key data memory 35B. The video data of the frame memory 35A is repeatedly read and output to the encoder 38 at the subsequent stage. The extracted data from the key data memory 35B is input to a built-in key circuit, and is output from the key circuit to the key output unit 39 as key data.

The timing generator 33 outputs the main image A from the pre-processing unit 31 so that the sub-image stored in the key data memory 35B can be read out in synchronization with the main image A.
A timing signal for memory reading is generated based on the synchronizing signal, and is sent to the memory controller 36.

FIG. 5 shows a detailed configuration example of a key circuit portion provided in the memory controller 36.
This key circuit receives input data (alpha data), color data (chroma key data), and luminance data (luminance key data), and outputs key data created based on these input signals. You. 361
Is a selector. The selector 361 receives input of extraction data, color data, and luminance data. The output signal selected by the selector 361 is a comparator 362.
Is output to the selector 367. This selector 361
When the title is selected as the sub-video, the selected data is selected and output to the selector 367. When the input video B is selected, the luminance data or the color data is selected and the comparator 362 is selected.
Output to The comparator 362 binarizes the input signal (luminance data) by comparing it with a threshold value, and
Selector 364 via a circuit 364 and an exclusive OR circuit 365
67 is output. The selector 367 selects an output signal from the selector 361 when a title is selected as a sub video, and selects an output signal from the exclusive OR circuit 365 when a input video B is selected.

The comparator 363 is used for wiping processing. The comparator 363 compares the wipe data from a wipe memory (not shown) with a threshold value and binarizes the output signal to AN.
Output to the D circuit 364. The exclusive OR circuit 365 is for controlling white / black, and receives a white signal or a black signal.

When the input image B (title original image) is selected as the sub image, the key circuit selects the luminance data in the image data by the selector 361 and inputs the luminance data to the comparator 362. Is compared with a threshold value to binarize, thereby extracting a character portion, and an AND circuit 364 and an exclusive OR circuit 36
5. The key data is output to the key output unit 39 via the selector 367 as key data. This key data is a signal in which the inside portion of the outline of the title character is "1" and the outside portion (that is, the background portion of the document) is "0". Note that the above-described threshold can be appropriately adjusted with the value key 251.

When a title is selected as a sub-picture, the key circuit operates in a key data memory 35.
Data is input from B. As will be described later, the extracted data is basically a signal in which the inner part of the outline of the character is binarized to “1” and the outer part (that is, the background part of the document) is “0” as described above. The extracted data is output to the key output unit 39 as key data via the selector 361 and the selector 367.

The key output section 39 creates an analog key signal based on the key data, and outputs it to the multiplier 44. As described above, the key data input to the key output unit 39 is for extracting the outline of the title (character). In this example, the inner side is 1 along the outline of the title and the outer side is along the outline. Is set to 0. The key output unit 39 converts the key data into an analog key signal synchronized with the video signal, and outputs the analog key signal to the multiplier 44. That is, the key signal can be said to be a video signal obtained by extracting a character portion (or a frame portion surrounding the character) from a video image of the character.

The encoder 38 comprises a D / A converter,
The video data is converted into an analog video signal by, for example, converting the color information into an analog color signal based on the input video data, and is transmitted to the input terminal of the modulator 42.

The modulators 41, 42, and 43 have the same configuration, and each receives a video signal.
It has two input terminals, and mixes and switches two input video signals according to a control voltage input to the control terminal. In this embodiment, the modulator 4
1, 42 and 43 output the video signal input to the input terminal when the control voltage is 0, output the video signal input to the input terminal when the control voltage is the maximum value, Output a video signal mixed with On this occasion,
If the control voltage is gradually changed between 0 and the maximum value, video switching is performed by cross-fade, and if the control voltage is changed instantaneously, normal video switching is performed.

The back color generator 37 is for generating a back color to be inserted at the switching point when switching images, and the back color image signal is input to the input terminals of the modulators 41 and 42. . The output signal of the modulator 41 is input to the input terminal of the modulator 43, and the output signal of the modulator 42 is input to the input terminal of the modulator 43. Modulator 41,
A control voltage is input from the other channel D / A converter 40 to 42, and a control voltage is input from the other channel D / A converter to the modulator 43 via the multiplier 44. A key signal is input from the key output unit 39 to the other input terminal of the multiplier 44. Instruction data (digital values) of these control voltages are supplied from the CPU 1 to the other channel D / A converter 40 through the interface 7.

The output signal of the modulator 43 is output as a video output to the monitor television 63 via the video encoder 45 and the video buffer amplifier 46. Here, the video encoder 45 adds the synchronization signal separated by the preprocessing unit 31 to the output video signal again.

FIG. 5 shows a detailed configuration of an audio circuit portion for performing audio processing in the video / audio circuit 6. This audio circuit comprises two circuits for the right channel and the left channel. The audio circuit receives an audio signal a for video A, an audio signal b for video B, and an external audio signal AUX such as BGM that is externally input regardless of the video A and B. A signal obtained by selecting or mixing an audio signal is output as an audio output.

Referring to FIG. 5, VCA (Voltage Controlled Amplifier)
Reference numerals 51 to 53 denote audio signals a, b, and A input to each of them.
This is a circuit for adjusting the volume of the UX according to the control voltage from the other channel D / A converter 56 and outputting the adjusted signal. The output signals are input to the adder 54. The adder 54 is a circuit for mixing the input signals, and the output signal is further output as an audio output via the VCA 55. This V
The CA 55 adjusts the volume level of the audio output according to the control voltage from the other channel D / A converter 56. In addition, instruction data (digital value) of a control voltage for each of the VCAs 51 to 53 and 55 is input from the CPU 1 to the other channel D / A converter 56 via the interface 7.

Hereinafter, various title processing methods which can be performed by the apparatus of this embodiment will be described. First, the basic operation of the video circuit shown in FIG. 4 will be described. here,
As shown in FIG. 7 (1), it is assumed that a video A in which a mountain is actually photographed is input to the video circuit from the video deck 61, and as shown in FIG. It is assumed that the image B, which is a real image of the title original image on which the character “mountain” is written, is input to the image circuit. This "mountain" character has a pattern. This allows
On the monitor television 63, an image in which the title character "mountain" is combined with the actual photographed image A of the mountain is displayed in a form according to various title processes described later.

As for the control voltage for each of the modulators 41 to 43, the modulator 41 always outputs only the input side image A, and the modulator 42 always outputs only the input side image B, in order to simplify the explanation. Thus, the control voltage (maximum value voltage) is set as described above, so that the input video A and the input video B are always input to the modulator 43.

The control voltage to the modulator 43 is at the maximum value level at the output of the other channel D / A converter 40. As described above, the key output unit 39 outputs the multiplier 4
The key signal input to 4 is “1” inside the outline of the character,
Since the outside is "0", the output of the multiplier 44 becomes zero in the outside part of the character, and this is input to the modulator 43 as a control voltage. Modulator 41
Output the input video A input to the input terminal. On the other hand, the inside portion of the character outline is obtained by multiplying the maximum value voltage from the other channel D / A converter 40 by the "1" signal from the key output unit 39 by the multiplier 44, and providing the modulator 43 with the maximum value voltage. Since it is input as a control voltage,
The modulator 43 outputs the image B (title image) input from the modulator 42 to the input terminal in the portion inside the character outline.

As described above, the image A input from the video deck 61 is transmitted through the modulator 41 to the modulator 4.
The video B (title image) input from the video camera 62 is input to the modulator 4 via the memory controller 36, the encoder 38, and the modulator 42.
3 is input. As described above, the modulator 43 outputs the image B from the modulator 42 in a portion corresponding to the character portion of the title original image, and outputs the image A from the modulator 41 in a portion other than the character (base portion). Accordingly, the title character is combined with the video A, and the resultant is output to the monitor television 63, and the video is displayed. Therefore, while watching the image on the monitor television 63, the threshold value for the key circuit is changed with the value key 251 so that the title image on the monitor television is adjusted to be the most beautiful.

Next, the outline of the processing in the CPU of this embodiment will be described with reference to the flowchart of FIG. In the apparatus of this embodiment, any one of the title processes I to V can be arbitrarily selected and performed (operators to be selected are not shown). The contents will be described later. First, a schematic processing operation in the CPU will be described with reference to the flowchart in FIG.

Now, it is assumed that a title original as shown in FIG. The background is shown in FIG.
It is assumed that a mountain landscape image as shown in (1) is input as image A. When the capture button 256 is pressed (step C11), the processing in FIG. 8 is started, the video data of the video B from the video camera 62 is captured by the frame memory 35A via the memory controller 36 for one frame, and The captured video (title video) is transferred to the RAM 2 (step C
12), stored as character surface data.

Next, it is monitored whether or not the character extraction switch button 257 has been pressed (step C13). If the button has been pressed, the binary data is compared by comparing the luminance data in the character surface data with a predetermined threshold value. And develops it on the RAM 2 as binary data (step C14). Although not shown, other processing may be performed when the character extraction switch button 257 is not pressed.

Next, it is detected whether or not edge processing is selected by a switch (not shown) (step C1).
5) If edge processing has been selected, the edge processing is performed (step C16). This edge processing will be described in detail later, but is described in Title Processing II and Title Processing III.
This processing is performed when the title processing VI is performed. When the title processing I and the title processing V are selected, the edge processing is skipped. The edge processing is performed on the binarized data in the title processing II and the title processing VI, and is performed on the character surface data stored in the RAM 2 in the title processing III.

In the title processing I and the title processing III, the binarized data not subjected to the edge processing is used. In the title processing II and the title processing VI, the binary data subjected to the edge processing is used.
In the title processing V, key-level data described later is stored in the memory 35B as extracted data.
(Step C17).

Next, when an instruction to save the data is given, the character plane data and the extracted data are stored in the respective title processing I
ＶV, the data is compressed by the following compression method (step C19). Then, each of the compressed data is transferred to and stored in the memory 9. (Title processing I, II, III) Character surface data → JPEG method Extract data → Run length method (Title processing VI) Character surface data → Run length method Extract data → Run length method (Title processing V) Character surface data → JPEG method Extracted data → key level data

Hereinafter, details of the title processes I to V performed by the apparatus of this embodiment will be described with reference to FIGS.

[Title Processing I] FIG. 9 shows the concept of the title processing I. FIG. 10 shows a flow of the title creation processing in the title processing I, and FIG. 11 shows a flow of the title display processing in the title processing I. In this title processing I, the title original image is
The data is compressed by an encoding method and stored as character surface data (frame data), and the character outline information of the title original image is compressed by a run-length encoding method and stored as extracted data. The original title character is restored as an output video based on the extracted data.

First, the title creation processing will be described with reference to the flow of FIG. As described above, the capture button 256 is pressed while the video in which the title character is inserted into the video A is displayed on the monitor television 63 (Step S).
1). Thus, the video data of the video B from the video camera 62 is captured by the frame memory 35A for one frame via the memory controller 36 (step S11).
Is transferred as character surface data (step S12).

Next, when the character extraction button 257 is pressed (step S2), the character surface data (the image data of the image B including the character surface pattern) is stored in the RAM2.
The middle luminance data is compared with a predetermined threshold value and binarized (step S13), thereby generating extracted data (step S14). The extracted data is data in which the inside portion of the character outline is “1” and the outside portion is “0”.

When the generation of the extracted data is completed, the display 4
Is displayed, "Do you want to register?" On the other hand, when the Yes button 254 is pressed (step S3), R
The extracted data generated on the AM2 is data-compressed by a run-length encoding method (step S15). At the same time, the character plane data (video B video data) on the RAM 2 is compressed using the JPEG encoding method (step S17).

Further, by pressing the storage bank number (title number) in the title / still image file memory 9 for storing this title using the title / still image file button 22, the corresponding storage bank number The extracted data (compression) by run-length compression and the character plane data (compression) by JPEG compression are stored in the storage area corresponding to (steps S16 and S18).

Next, the title display processing will be described with reference to the flow of FIG. The title number of the title image stored in the title / still image file memory 9 as described above is designated by the title / still image button 22, and then the title button 242 is pressed (step S51). As a result, the title image data [extracted data (compressed) and character plane data (compressed)] of the designated title number is read from the title / still image file memory 9. The extracted data (compression) in the title image data is expanded according to the run-length encoding method and restored to the original binary data (step S53), and the key data memory 3
5B and stored (step S545). Character data (compression) in title image data is JP
The data is decompressed according to the EG method, restored to the original character surface data, transferred to the frame memory 35A and stored.

As described above, as described above, only the character portion in the title original image is restored, and is displayed on the image A by being imposed (step S57). According to the method of title processing I, even if a character decompressed from character surface data (compression) has a so-called "whisker" in its outline, this "whisker" is located outside the title outline specified by the extracted data. Thus, the title image is removed when the title image is restored, and the quality of the restored title character image is improved.

[Title Processing II] FIG. 12 shows the concept of title processing II. FIG. 13 shows title processing II.
2 shows the flow of the title creation processing in FIG. The flow of the title display process in the title process II is the same as that in FIG. 11 described above. The basic principle of this title processing II is the same as that of the above-mentioned title processing I, but edge processing is performed on the extracted data so that the outline of the character becomes beautiful and familiar with the background image. ing.

In FIG. 13, the difference from the title creation processing in the title processing I shown in FIG.
After binarizing the luminance data on the RAM 2 (step S
13) In order to use this as extracted data, the binarized data is subjected to edge processing (step S).
14 '). The edge processing is processing for inserting halftone (halftone) data inside the outline (that is, edge) of the character. FIG. 14 is a diagram for explaining this edge processing. After the captured luminance data is binarized, “1” (white frame outline of the character outline) contacts the “0” (black frame outline of the character outline). Pixels inside are replaced with halftone “0.5” (hatched frame) data.

After generating such extracted data, the extracted data is compressed by a run-length encoding method and stored. When a title is displayed, the extracted data is reproduced to generate a key signal, and a multiplier 44 in the video circuit is used.
, The control voltage input to the modulator 43 is multiplied by 0.5 by the multiplier 44 in this halftone portion, and thus becomes half of the maximum value voltage. Therefore, the modulator 43 outputs an image in which the background image A and the character image B are mixed in the halftone portion, so that the outline portion of the title character becomes the background image A.
Will be projected with a familiar feeling.

This title processing II can be modified as follows by further adding attribute information. 15 and 16 show this modification. FIG. 15 shows a flow of a modified example (registration of attributes) of the title reproduction process in the title process II, and FIG. 16 shows a flow of a modified example (use of attributes) of the title display process.

The modification shown in FIG. 15 is different from the title creation processing shown in FIG. 13 in that the Yes button 254 is displayed in response to the “register?”
After pressing (step S3), guidance of "designation of how to put out?" Is displayed (step S5). In response to the guidance, mix, wipe (type), scroll (direction), and speed are displayed. It can be specified.
The specified data is stored as attribute data in the title / still image file memory 9 together with the title image data (step S19).

When displaying the title image data, as shown in FIG. 16, this attribute data is also read from the title / still image file memory 9 (step S58), and when the title-in button 267 is pressed (step S58). In step S60, the control data specifying data based on the attribute data is set in the other channel D / A converter 40 of the video circuit (step S59), and the title character is set to the mix, wipe type, and scroll direction specified by the attribute data. Insert title characters by speed, speed, etc.

[Title Processing III] FIG. 17 shows the concept of title processing III. FIG. 18 shows the flow of a title creation process in the title process III. The flow of the title display processing in the title processing III is substantially the same as the flow of the title display processing in the title processing I shown in FIG. The title processing III has the same basic principle as that of the above-described title processing I. However, in order to clearly distinguish the restored title character from the background image (input image A), the outline of the character surface is added. It is bordered along.

In FIG. 18, the flow of the title creation processing in the title processing III is almost the same as the flow of the title creation processing of the title processing I shown in FIG. The difference is that the character plane data transferred to the RAM 2 (step S12) is compressed by the JPEG encoding method (step S17).
The point is that a process for bordering the outline of the character surface is performed based on the binarized luminance data on the RAM 2 (step S13) (step S20). This bordering is performed by selecting a border of a color different from that of the background video (input video A), a width of the border that is easy to see, and the like.

The flow of the title display process in the title process III is the same as the title process I shown in FIG.
Since the title display process is substantially the same as that of the title display process described above, a detailed description thereof will be omitted. However, since the title characters to be restored are bordered, the characters are conspicuous with respect to the background video.

[Title Processing VI] FIG. 19 shows the concept of the title processing VI. FIG. 20 shows a title processing VI.
FIG. 21 shows the flow of a title creation process in the title processing VI. In the title processes I to III described above, frame data (character surface data) is subjected to data compression by the JPEG encoding method. However, in the title process VI, the character surface is color-processed into a pattern. Is different from the title processes I to III in that data is compressed by the run-length encoding method as frame data.

In the title creation processing shown in FIG. 20, after character extraction (step S2), the color processing button 25
Press 8 (step S6). Thus, it is possible to specify various color processing. Specifically, this color processing is performed, as shown in FIG. 19, by converting a character surface having a pattern in a manuscript image into a solid character, bordering the outline thereof, and further adding a shading to the character pattern. It is assumed that. Therefore, as the specification of the color processing, it is possible to specify a character color, a border color, a border width, a shade color, a shade width, a shade direction, transparency, and the like.

This color processing is performed in RAM 2
This is performed on the coded luminance data (step S2
1). Further, edge processing is performed on the color-processed character surface data to create cutout data (step S2).
2). Then, in response to the "Do you want to register?"
When the s button 254 is pressed, the extracted data is encoded by the run-length encoding method (step S1).
5) The frame data (character surface data) subjected to the color processing is also encoded by the run-length encoding method (step S17 '). The extracted data (compressed) and the frame data (compressed) are used as title image data for the title /
The image data is stored in the still image file memory 9 (step S1).
6, S18). In the title creation processing by the title processing VI, the encoding of the designed title characters does not have a large data amount, so that the amount of stored data can be reduced.

The flow of the title display process in the title process VI is substantially the same as the title display process of the title process I shown in FIG.

[Title Processing V] FIG. 22 shows the concept of title processing V. FIG. 23 shows title processing V
24 shows a flow of a title creation process in FIG. 24, and FIG. 24 shows a flow of a title display process in the title process V. In the title processing V, the data corresponding to the extracted data in the above-described title processing I to VI is realized and stored as key level data. The extracted data is compressed by run-length encoding and stored. I have not done it.

In the title creation processing shown in FIG. 23, first, a white / black button is pressed to specify the background color of the original from which the title is to be extracted (step S7).
As a result, the memory controller 36 enters the luminance key mode (step S24), and in the key circuit, the selector 361 selects luminance (luminance) data and supplies it to the comparator. At this time, the selector 367
Selects the output signal from the exclusive OR circuit 365 side.

The value button 251 is operated to adjust the threshold value (extraction level). As a result, the set value of the threshold value of the key circuit is sequentially changed, and a key output in which characters are extracted according to each set value is output (step S1).
2) Since the video of the extracted title is displayed on the monitor 63, the value is set so that the title video is visually excellent. After the setting is completed, the title mode (key mode of white or black and key level (value value)) used at that time is recognized (step S13), and the title / still image file memory is used as attribute data of the title. 9 (step S14).

Next, when the capture button 256 is pressed (step S3), the image B from the video camera is captured in the memory 35A (step S15). The image B in the memory 35A is further transferred to the RAM 2 and stored as character surface data.

Next, the display of “Do you want to register?” Is displayed on the display 4. If the Yes button 254 is pressed (step S 4), the image B stored in the RAM 2 is displayed.
Is encoded using the JPEG method, thereby compressing the character surface data (step S17). Then, the compressed data is stored in the title / still image file memory 9 as character surface data (frame data) (step S18), and the key mode used for extracting the title (extracted color), the key level, and the title mode ( Here, the luminance mode is stored in the title / still image file memory 9 as attribute data of the character surface data (step S27).

Next, the title display processing will be described with reference to the flow of FIG. When the title number is designated by the title / still image button 22 and the title button 242 is pressed (step S51). The title image data [key level data and character surface data (compressed)] of the designated title number is read. The character surface data (compression) in the title image data is expanded according to the JPEG method and restored to the original character surface data (step S53), and the frame memory 3
5A and stored (step S54). On the other hand, the key level data in the title image data is read out (step S61), and based on the key level data, the setting of the threshold value of the key circuit in the memory controller 36 and the setting of white / black are performed. , Setting of the luminance mode (setting of switching the selectors 361 and 367 to the computer 362 side) is performed (step S).
62). Thus, the title character is extracted from the restored video data set in the frame memory 35A and displayed on the monitor television 63.

As described above, since the attribute data is stored in the title image data in the title / still image file memory 9, the title character can be read by calling the attribute data without performing various adjustments of the key circuit at the time of restoration. Can be extracted and projected.

[Merging Processing] Next, merging processing between title / still picture files will be described.
FIG. 25 shows the flow of this merge processing.
FIG. 4 is a diagram illustrating the concept of various forms of merge processing.

When the merge button 264 is turned ON (step S71), a message "Please specify a file" is displayed on the display 4 (step S72). Specify using the title / still image file button 22 (step S8)
1). This is a 2 like "A1" and "A4"
Specify two files.

Then, a guidance "Please specify the type of merge" is displayed on the display 4 and the number is specified using the title / still image file button 22 (step S82). ). Here, there are three types of merge types, “1”, “2”, and “3”. The merge type number “1” is the merge process shown in FIG. 26, and the merge type number “2” is the merge process shown in FIG. , The merge type number “3” is the merge process shown in FIG. Then, a merge process corresponding to the selected merge type number is executed (step S83).

The merging process will be described below. In the merging process, when two screens are mixed, a portion having high luminance is preferentially blended, and a portion having low transparency is preferentially blended. And

The process of merging type number "1" is a process of merging a title with a still image as shown in FIG.
Are merged. This merging is performed by extracting the area indicated by the extracted data in the still image using the extracted title data, and inserting the frame data of the title therein.

As shown in FIG. 27, the process of the merge type number "2" is a process of merging titles by the "sum" logic. The title read as the file 2 is inserted into the title read as the file 1. To merge. In this merge, the area indicated by the extracted data of file 2 is extracted from the title of file 1 and the frame data of the title of file 2 is inserted therein.

As shown in FIG. 28, the process of the merge type number "3" is a process of merging titles by the "product" logic, and is performed between the title read as file 1 and the title read as file 2. The file 2 is merged in such a manner that the title face of the file 2 is inserted into the logical product. That is, this merge is performed for a portion where the title of file 1 and the title of file 2 overlap.
The frame data of the title of file 1 is inserted into the overlapping portion (product portion).

According to the merge type number "3", for example, the character pattern data for one screen is stored as the file 1, and the title is stored in the file 2, and the two are merged. In the character outline of the title of the file 2, the pattern of the title 1 can be attached as its character surface.

Various modifications are possible in practicing the present invention. In the above-described embodiment, the JPEG encoding method is used as the encoding method for compressing the title surface data (character surface data). However, the present invention is not limited to this, and various encoding methods that can expect high compression efficiency are used. Irreversible coding scheme can be used. This irreversible encoding means that even if the original video is not physically completely reproduced, it only needs to look the same as the original video when viewed by human eyes.
Data compression is performed by effectively removing visual redundancy using this human visual characteristic. Statistical redundancy compression such as orthogonal transformation (DCT, wavelet, Gabor function , Pyramid coding, fractal coding, etc.).

On the other hand, in the above-described embodiment, the run-length encoding method is used as the encoding method for compressing the extracted data, but the present invention is not limited to this.
Various lossless encoding schemes (or information storage encoding schemes) that can be expected to have high reproducibility can be used. The lossless encoding method is an encoding method for completely restoring the original video from the encoded data without distortion, and mainly uses direct encoding (for example, run-length encoding, Whle encoding, Shannon-Fano encoding, Huffman coding, extended Huffman coding (MH code), multiple line batch coding, etc.) are included in this.

Further, in the title processing II of the above-described embodiment, the edge-processed extracted data is run-length coded and stored. However, the present invention is not limited to this, and the binarized extracted data is run-length encoded. Encode and save,
When decompressing the extracted data, edge processing may be performed on the extracted data after decompression, and the title video may be restored using the extracted data subjected to the edge processing.

Similarly, the title processing III of the above embodiment
In the above, the data obtained by performing the border processing on the character surface data was JPEG-encoded and stored. However, the present invention is not limited to this, and the character surface data itself is JPEG-encoded and stored, and when decompressing it, The frame data may be subjected to edging processing, and the title video may be restored using the text data subjected to the edging processing.

In the above-described embodiment, the title picture is a character such as a title. However, the present invention is not limited to this. It is a concept that includes.

Further, in the present invention, a control program describing a procedure for performing the above-described title processes I to V can be stored in a computer-readable recording medium and distributed.

[0094]

As described above, according to the present invention, it is possible to improve the quality of a restored title video while reducing the amount of data stored in the title video.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a diagram schematically showing an overall configuration of an AV mixer equipped with a title video storage / playback apparatus as one embodiment of the present invention.

FIG. 3 is an external view of an operation panel in the embodiment device.

FIG. 4 is a diagram showing a detailed configuration of a video circuit portion related to video processing in a video / audio circuit of the embodiment device.

FIG. 5 is a diagram showing a detailed configuration of a key circuit portion provided in a memory controller of a video circuit of the embodiment device.

FIG. 6 is a diagram illustrating a detailed configuration of an audio circuit portion related to audio processing in a video / audio circuit of the embodiment device.

FIG. 7 is a diagram showing a video A and a video B input to the embodiment device.

FIG. 8 is a flowchart illustrating a schematic process executed by a CPU in the embodiment device.

FIG. 9 is a diagram showing the concept of title processing I.

FIG. 10 is a diagram showing a flow of title creation processing in title processing I.

FIG. 11 is a diagram showing a flow of title display processing in title processing I.

FIG. 12 is a diagram showing the concept of title processing II.

FIG. 13 is a diagram showing a flow of title creation processing in title processing II.

FIG. 14 is a diagram for explaining edge processing in title processing II.

FIG. 15 is a diagram showing the flow of another example (attribute registration) of title creation processing in title processing II.

FIG. 16 is a diagram showing a flow of another example of title display processing (use of attributes) in title processing II.

FIG. 17 is a diagram showing the concept of title processing III.

FIG. 18 is a diagram showing a flow of title creation processing in title processing III.

FIG. 19 is a diagram showing the concept of a title processing VI.

FIG. 20 is a diagram showing a flow of a title creation process in the title process VI.

FIG. 21 is a diagram showing a flow of a title display process in a title process VI.

FIG. 22 is a diagram showing the concept of title processing V.

FIG. 23 is a diagram showing a flow of title creation processing in title processing V.

FIG. 24 is a diagram showing a flow of a title display process in a title process V.

FIG. 25 is a diagram showing a flow of a merge process.

FIG. 26 is a diagram conceptually illustrating a merge process of a merge type number “1”.

FIG. 27 is a diagram conceptually illustrating a merge process of a merge type number “2”.

FIG. 28 is a diagram conceptually illustrating a merge process of a merge type number “3”.

[Explanation of symbols]

 1 CPU (Central Processing Unit) 2 RAM (Random Access Memory) 3 ROM (Read Only Memory) 4 Display 5 Display Driver 6 Video / Audio Circuit 7 Interface 8 Operators 9 Memory for Title / Still Image File Reference Signs List 20 switch button for video / audio switching 21 switch button for fade control 22 switch button for title / still image file 23 fader 24 switch button for sub-video 25 other switch buttons 31, 32 preprocessing unit 33 timing generator 34 A / D converter 35A Frame memory 35B Key data memory 36 Memory controller 37 Back color generator 38, 45 Encoder 39 Key output unit 40, 56 Other channel D / A converter 41-43 Modulator 44 Multiplier 4 Video buffer amplifier 51~53,55 VCA (voltage controlled amplifier) 54 adder 361,367 selector 362 comparator

Claims (5)

[Claims] 1. A storage means for compressing and storing a title image as surface data by irreversible encoding as title data, and compressing and storing the title outline as lossy data by lossless encoding. Restoring means for restoring a title outline from the compressed extracted data and restoring the title image by synthesizing a title plane restored from the compressed plane data within the title outline, apparatus. 2. When saving or restoring the extracted data, edge processing is performed to add a halftone edge along the title outline of the extracted data. 2. The title video storage / reproduction apparatus according to claim 1, wherein the video of the surface and the background video are mixed. 3. The title image storage and reproduction according to claim 1, wherein, when the plane data is stored or restored, a border attaching process for adding a border to the title plane of the plane data along the title outline is performed. apparatus. 4. A title image obtained by subjecting the title surface of the title image to a design process is subjected to data compression by lossless encoding as surface data and stored, and data compression is performed by lossless encoding of the title outline subjected to the design process as extracted data. Storing means for restoring the title; restoring a title outline from the compressed extracted data; and stitching the title by combining a stylized title surface restored from the compressed surface data within the title outline. An apparatus for storing and reproducing a title image, comprising: a restoration unit for restoring an image. 5. A title image is subjected to data compression by lossless encoding as title data, and is saved as lossy data, and a pattern image to be attached to the title surface is compressed by irreversible encoding and saved as surface data. Storage means, and restoration means for restoring a title outline from the compressed extracted data and restoring the title video by combining a pattern image restored from the compressed plane data within the title outline. Title video storage and playback device.