JP3477750B2 - Face image creation device and face image creation method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a face image forming apparatus and
A method for creating a face image .

[0002]

2. Description of the Related Art Conventionally, a graphic display device using, for example, a microcomputer is used to create a face image. In this device, each part (eg,
With parts images that represent each part ( eye, eyebrow, mouth, etc.)
A face image is created by combining certain part patterns. Specifically, hairstyle, face contour, eyes, eyebrows, nose,
Equipped with a part memory that stores each part of the mouth etc. as multiple kinds of bitmap patterns, and selectively call these individual part patterns by key input operation to synthesize and display on the display unit and create any portrait image is doing.

[0003]

However, when the face image is created by the conventional graphic display device, if there are few kinds of part patterns stored in the parts memory in advance,
There is a problem that it is not possible to create a caricature that approximates a real face . Therefore, for example, it is also possible to create a caricature similar to the actual face by giving many types of patterns for each part in the above-mentioned part memory, but the types of part patterns are too large. As the number of parts increases, it takes a lot of time to select each part pattern, and it is difficult to determine one of a plurality of approximated part patterns as an appropriate part pattern.

On the other hand, the applicant of the present invention has previously proposed a technique for judging one of a plurality of approximated part patterns as an appropriate part pattern. In this technique, each partial image of the input image and By comparing multiple types of part patterns for each part stored in the parts memory in advance, and selecting and combining the part patterns for each part with the highest pattern matching degree, for example, the actual face I try to create a portrait similar to <br/>. However, even with this method, the position data of the part pattern is not used at all, and it is not possible to correct the fine position, and it is not possible to create a caricature that more closely resembles the actual face . If this method is used ,
If you try to create a caricature that approximates a face , you also need part patterns that have the same shape but different positions, and you need to prepare an extremely large number of part patterns in advance. It takes an enormous amount of labor and time to select, which is not realistic.

Therefore, according to the present invention, it is possible to easily create a face image that is very similar to the acquired face image by combining the part images, without the trouble of selecting the part pattern. Device and facial image creation method
It is intended to provide the law .

[0006]

To achieve the above object, a face image forming apparatus according to the invention of claim 1 comprises the following means.
It is characterized by that.

That is, the face image creating apparatus according to the invention of claim 1 extracts a partial image for each part of the face from the photographed face image prior to the reproduction of the face image, and the extracted partial image and face A plurality of types of part images stored in advance for each part are compared for each part, and based on the comparison result, a part image corresponding to the part image for each part stored in advance is designated. A storage control unit that associates the data with the coordinate position data for each part of the partial image, and sequentially stores the data in the storage unit as time passes,
During reproduction of the face image, the part image corresponding to each stored part image designation data and the coordinate position data stored in association with each part image designation data are stored in the storage unit as time passes. Sequential reading,
The parts images are arranged at positions corresponding to the read out parts image designation data, and the portrait images created by combining the arranged part images of the respective parts are sequentially displayed on the display unit. And a display control means.

[0008]

[0009]

[0010]

[Action] In the face image creation device according to a first aspect of the present invention, the following
To act.

That is, in the face image creating apparatus according to the first aspect of the present invention, the storage control means extracts the partial image for each part of the face from the photographed face image prior to the reproduction of the face image, and this extraction is performed. The partial image and a plurality of types of parts images stored in advance for each part of the face are compared for each part, and based on the comparison result, the parts image for each part stored in advance is Corresponding parts image designation data and coordinate position data for each part of the partial image are associated with each other and sequentially stored in the storage unit as time passes. The display control means, at the time of reproducing the face image, the part image corresponding to each stored part image designation data and the coordinate position data stored in association with the respective part image designation data from the storage unit. Are sequentially read out with the passage of time, the parts images are arranged at the positions corresponding to the respective parts image designation data read out, and the portraits created by combining the parts images of the arranged parts The images are sequentially displayed on the display unit. Also, in the face image creating method according to the invention of claim 3,
The same operation as in the face image creating apparatus according to the first aspect of the invention is performed.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a face image creating apparatus according to the present invention and
It is a figure which shows 1st Example of the face image creation method , and is a front view of an electronic notebook especially when this invention is applied to an electronic notebook. In FIG. 1, the electronic notebook 1 is a pair of main bodies 1
It shows a state in which a and 1b are respectively opened to the left and right. The CCD camera unit ( face image) is provided on top of the main body 1b of the electronic notebook 1.
(Acquisition means ) 2 is arranged. The CCD camera unit 2 is directed to the face of an operator (operator) existing almost in front of the electronic notebook 1, and the operator's face is imaged to generate image data, and the image data is read to read the main body 1a,
A Y / C separation unit 23a shown in FIG.
Output to. That is, the CCD camera unit 2 has an image pickup unit and a data unit that processes a captured image and reads it as image data. As a face image acquisition means ,
Not limited to the CCD camera, for example, a video camera, an image scanner, an electronic still camera or the like may be used. Alternatively, an image signal from another device may be used.

A second display portion 3 and a data input key 4 are sequentially arranged below the CCD camera portion 2 in the main body 1b of the electronic notebook 1. The second display unit 3 is composed of, for example, an LCD (liquid crystal display device), and displays the face image 2a captured by the CCD camera unit 2, so that the captured face image 2a can be displayed as it is. Has become.
The second display unit 3 is an LCD capable of displaying a monochrome screen.
Alternatively, it may be an LCD capable of displaying a color screen. The data input key 4 is for inputting various data necessary for operating the electronic notebook 1, for example, "1",
Numeric key consisting of "2", ... "9", "0",
Operation keys such as "-", "+", ..., "A",
It has alphabetic keys consisting of "B", ... "Y", "Z". By using these numerical keys, calculation keys, and alphabet keys, various data about a person who is a target person of face image creation is input, or data other than the image is input.

Below the main body 1b of the electronic notebook 1, a printing unit (see FIG. 2 described later) capable of printing the label tape 4 is built in. In the example of FIG. 1, the face image displayed on the first display unit 11 is displayed. The state where 11a is printed on the label tape 4 and taken out is illustrated. The label tape 4 is attached to the label tape 4 on which a face image is printed at an appropriate position by the adhesive action of the adhesive layer by peeling off the release sheet provided on the back side of the label tape so as to be removable via the adhesive layer. It is possible. Further, not only an image but also a sentence can be printed on the label tape 4. It should be noted that not only the face image 11a displayed on the first display unit 11 but also an image displayed on the second display unit 3 described later can be printed on the label tape 4 and taken out. The printing material may be plain paper instead of label tape.

On the other hand, the main body 1a of the electronic notebook 1 is provided with a first display portion 11, an input switch portion 12, an on / off key 13, and a site key 14 from the upper side. First display unit 1
Reference numeral 1 denotes, for example, an LCD, which displays various images and a composite face image at the stage of electronically creating a face image, and displays necessary data (for example, data such as a person's name). The first display unit 11 may be an LCD capable of displaying a monochrome screen or an LCD capable of displaying a color screen. In the example of FIG. 1, a face image 11 of a person
The name, address, telephone number, and age are displayed as data relating to a and the person. Of course, data other than these may be used. Input switch unit 12
Is for performing various operations necessary for creating a face image of a person using the electronic notebook 1, and specifically, the following keys are provided. For the following keys, for example, thin sheet type key switches are used. "Matching": When matching is performed by comparing a partial image corresponding to a partial image of the face stored as the read data of the CCD camera unit 2 with a plurality of types of part patterns for each stored face part. Key to operate. “Display”: a key operated when displaying an image on the first display unit 11. Alternatively, the operation may be performed when the image is displayed on the second display unit 3. “Print”: A key operated when printing an image or a sentence on the label tape 4.

"Individual": A key operated when displaying data relating to an individual, for example, data relating to an individual such as name, address, telephone number, etc., on the first display section 11. “Enlargement”: A key operated when enlarging the part of the face image displayed on the first display unit 11. "Image input": Face image 2a captured by the CCD camera unit 2
Key to operate when fixing and displaying. “Position correction”: a key operated when correcting the position of the part of the face image displayed on the first display unit 11. “Correction”: This is operated when correcting the face image displayed on the first display unit 11. "△ ▽": A selection key operated when selecting a desired part pattern from face parts or multiple parts,
A so-called increment / decrement operation. △ indicates up, and ▽ indicates down. “End”: A key operated when ending the processing of the face image displayed on the first display unit 11. “Register”: Operates when registering the face image displayed on the first display unit 11. “Face direction”: A key operated when changing the direction of the face image displayed on the first display unit 11 (for example, displaying a side face). The on / off key 13 is operated when the power of the electronic notebook 1 is turned on / off. Part key 14
Is operated when designating each part of the face (for example, hair, eyes, nose, mouth, etc.), and a desired part is designated by pressing a graphic type key imitating each part. It should be noted that the body, hands, feet, etc. of the person are merely displayed graphically, and the function as a switch is omitted.

Next, FIG. 2 is a block diagram showing the configuration of the face image creating apparatus in the electronic notebook. In FIG. 2, the face image creating apparatus includes a CPU 21, a key input unit 22, an image data input unit 23, a pattern matching unit 24, a print drive unit 25, a print unit 26, a display drive unit 27, an input image memory 28, and a basic position memory. 29, parts pattern RO
It is configured to include the M30, the ROM 31, the work RAM 32, and the first display unit 11 and the second display unit 3 described above. The key input unit 22 includes the data input key 4, the input switch unit 12, the on / off key 13 and the part key 14 shown in FIG. 1, and the key operation signal from the key input unit 22 is input to the CPU 21. CPU 21 is the key input unit 2
The CPU 21 controls the operation of each part of the circuit in accordance with the portrait creation program stored in the ROM 31 based on the key operation signal from the key input unit 22 and the ROM.
In addition to 31, the image data input unit 23, the pattern matching unit 24, the print driving unit 25, the display driving unit 27, the input image memory 28, the basic position memory 29, the part pattern ROM 30, the ROM 31, and the work RAM 32 are connected. CPU21, together with ROM31, comparison
Means 101, creating means 102, and time measuring means 103 are configured. In addition, in order to correspond to the invention of claim 2 and below ,
CPU21, together with ROM31, extraction means, selection hand
The functions as a stage, a display position adjusting means, and a display control means are realized.

The image data input unit 23 is a unit for inputting a face image (for example, the face image 2a shown in FIG. 1) of a person who is a portrait image taken by the CCD camera unit 2 as a video signal (for one field). The image data input unit 23 includes a Y / C separation unit 23a and an A / D conversion unit 23b. The Y / C separation unit 23a separates and extracts only the luminance signal Y from the luminance signal Y and the color difference signal C included in the video signal from the CCD camera unit 2, and the video signal from the Y / C separation unit 23a is A / C. It is supplied to the D conversion unit 23b. The A / D conversion unit 23b converts the video luminance signal Y for one field supplied from the Y / C separation unit 23a into a monochrome binary digital signal in synchronization with a preset sampling clock. The image data converted into digital data via the D conversion unit 23b is transferred to and stored in the input image memory 28. That is, the input image memory (face image data storage means) 28 stores face image data taken as a portrait portrait person as bitmap data. The face image data stored in the input image memory 28 is stored by setting the image area in which the "nose" of the face exists in the center of the storage area.

The basic position memory 29 is the input image memory 28.
The basic memory locations of face images in hair, eyebrows, eyes, nose,
The facial image data stored in the input image memory 28 according to the address value stored in the basic position memory 29 is stored as the address value divided for each part such as the mouth and the ear. The face image data captured as is read out for each part such as hair, eyebrows, eyes, nose, mouth, and ears. Parts pattern ROM ( parts storage means, first
(2 storage means ) 30 includes various parts of the face such as hair, eyebrows, eyes, nose, mouth,
It divides into parts such as ears, and stores a plurality of types of part patterns for each part as bitmap data by assigning numbers to predetermined storage positions, and the part patterns for each part stored in the parts pattern ROM 30 are respectively stored. A portrait image is constructed by combining them one by one.

Here, an example of a part pattern of each part of the face stored in the part pattern ROM 30 is shown in FIG. In FIG. 3, as each part of the face,
A plurality of part patterns are set corresponding to the orientation of the face, such as front (A), right side (B), ... diagonal right side (N). And for each surface (A)
In (N), each part of the face is divided into parts patterns such as the outline of the face, the hairstyle, the right eye, ... The face part is not limited to the example shown in FIG. 3, but other parts such as eyebrows and nose are prepared, though not shown. Further, other face parts may be prepared.

On the other hand, as a part pattern of each part of the face, a plurality of kinds (here, 50 kinds) indicated by part pattern numbers [01], [02], ... [50] are provided in advance. , Contour, hairstyle, right eye, ... are prepared and stored for both feet. Further, the position data is attached as coordinates [x, y] to each part pattern of each part,
It is stored in a predetermined area. Therefore, each part has a plurality of part patterns together with position data of each part pattern. Further, in the part pattern ROM 30, the number of matching matching dots with the face image data corresponding portion stored in the input image memory 28 corresponding to each of the part patterns [01] to [50] provided in plural for each part. A match frequency register that stores
There is provided a rank register in which rank numbers are stored in descending order of the number of matching dots stored in the matching frequency register. FIG. 4 shows, as an example, a stored state of a part pattern corresponding to one of the face parts, that is, a "nose", in particular, the number of matching matching dots and the order number.

The work RAM ( first storage means ) 32 extracts a partial image for each part of the face from the face image data stored in the input image memory 28, the orientation data of the extracted partial image, and each part Store position data. The pattern matching unit 24 stores the part data corresponding to the partial image stored in the work RAM 32 and the part pattern ROM 31.
This is to sequentially compare (comparate) with a plurality of types of part patterns for each part of the face stored in, and find the number of matching dots corresponding to the number of image matching points. The number of matching matching dots is stored in the matching frequency register described above, and
The rank of the matching frequency is stored in the above-mentioned rank register.

When the number of matching dots for each part of the face image data and its rank number are stored for all of the plurality of part patterns for each part in the part pattern ROM 30, the first rank is set for each part. The part patterns of are read out and combined and displayed corresponding to each position data. The display drive unit 27 drives the first display unit 11 and the second display unit 3 based on the output of the CPU 21 to display the processed image. Further, the print drive unit 25 (output means) drives the print unit 26 based on the output of the CPU 21, and the print unit 26 prints a processed image, characters, etc. on the label tape 4. The display drive unit 27 and the first display unit 11 form a display unit.

Next, the operation will be described. First, electronic notebook 1
When the main bodies 1a and 1b are opened to the left and right and the on / off key 13 is pressed, power is supplied to the electronic notebook 1 and the operation starts. After that, by executing the program described later, the personal data is input and the portrait creation process is performed. FIG. 5 is a flow chart showing a processing program when inputting / registering personal data using the electronic notebook 1 of this embodiment. When this program is started, it is first determined in step S10 whether or not the private key is pressed. The personal key is an “individual” key in the input switch unit 12, and is data relating to an individual, such as a name,
Data about an individual such as an address and a telephone number is displayed on the first display unit 1
It is operated when it is displayed on 1. If the individual key has not been pressed, the process stays in step S10. If the individual key has been pressed, the process proceeds to step S12, and it is determined whether or not the individual data has been input.

The personal data is input by using the data input key 4, and it is possible to input personal data such as name, address and telephone number. In this case, if personal data has not been input, the process jumps to step S16. If the personal data is input, the process proceeds to step S14, and the input personal data is stored in the work RAM 3
Store in 2. As a result, the CPU 21 causes the work RAM to
The input personal data stored in 32 is read and output to the display drive unit 27, and the display drive unit 27 drives the first display unit 11, whereby the input personal data is displayed on the first display unit 11. . Next, in step S16, it is determined whether or not the registration key is turned on. If it is not turned on, the process returns to step S12 and steps S12 to S1.
Repeat the loop of 6. When the registration key is turned on in step S16, the input personal data is registered and the present routine is terminated. As a result, the personal data is stored in the predetermined registration area of the work RAM 32, and can be subsequently taken out as registered personal data.

FIG. 6 and FIG. 7 are flowcharts showing a program for portrait creation processing. When this program starts, it is first determined in step S100 whether the display switch is on. The display switch is a display key in the input switch unit 12, and when the display switch is turned on, the first display unit 11 is activated to display an image, and the process proceeds to step S102 and is already created. A process of selecting and displaying a portrait is performed. If the display switch is not turned on, the process branches to step S104, and the process of creating a portrait from an actual person image is executed. A. Routine of portrait selection processing First, the routine of portrait selection processing will be described. In step S102, it is determined whether search data has been input. The search data is data for selecting the created portrait and is, for example, personal data. If search data has not been input, the process returns to step S100. If search data has been input, the process proceeds to step S106 and the corresponding data is searched from the personal data in the work RAM 32 by the search data and displayed on the first display unit 11. indicate.

Here, FIG. 8 shows an example of the storage of the work RAM 32. For example, the facial features of each person correspond to the personal data of the person, the direction of the face, the part pattern number of each part of the face, The position data (parts data) of each part is stored in a predetermined area as a parameter.
Also, in this case, the face orientation (ie front, right side,
.... Contour, hair, right eye, ...
Part pattern numbers such as, for example, [50],
[02], [50], ... are stored in a predetermined area, and the outline, hair, right eye, ...
・ Position data of parts such as, for example, [x1, y
1], [x2, y2], [x3, y3], ..., And stored in a predetermined area. [X1,
The values y1], [x2, y2], [x3, y3], ... Are coordinate values that specify the position on the screen, and are difference data from the absolute position according to the position correction. . If the position is not corrected, the coordinate value is [0,0].

Then, for example, a part pattern number for each part and a position data of each part of the face image corresponding to the orientation of the face (that is, the front, right side, ..., oblique right side) are combined and combined. By doing 1
A montage image that is one portrait image is created.
Therefore, when the input search data is personal data, the corresponding personal data is collated, and the matching personal data is selected and displayed on the first display unit 11. For example, if any of personal name: former address: ○○ telephone number: ○○ blood type: ○ hobby: ○○ is input, it is stored in the first area of FIG. 8 as the corresponding person. The personal data of the person being selected is selected and displayed on the first display unit 11. As personal data, in addition to the example of the person A, the following person data is stored. Further, other person data may be stored. Name: B Address: ○○ Phone number: ○○ Blood type: ○ Hobbies: ○○

Next, one face image data in the area of the work RAM 32 in which the montage image (that is, the portrait image ) corresponding to the personal data retrieved in step S108 is stored is read out on the first display section 11. indicate. As a result, one of the portraits already created based on the personal data is picked up and displayed. For example, when the data relating to the human instep as described above is input as personal data, the orientation of the face corresponding to the instep shown in FIG. 8, the part pattern number for each part of the face, and the position of each part 1 based on data (parts data)
Two caricatures are displayed. Therefore, although there are a plurality of portraits corresponding to the instep of the person depending on the direction of the face, etc., one of them is selected and displayed.

Then, in step S110, the selection key (Δ
Determine whether or not the ▽ key) is turned on. If the selection key is on, it is determined that another portrait is requested even if the personal data is the same, and in step S112 the work RAM is selected.
Other face image data in the area 32 is selected. Then, the process returns to step S108 and the same loop is repeated. Thereby, for example, when the data related to the person A is input as the personal data as described above, another portrait corresponding to the person A is selected and the first display unit 11 is selected.
Is displayed in.

At this time, if the desired portrait corresponding to the retrieval data of the person's instep is not displayed on the first display section 11, the selection key (Δ ▽ key) is further operated (for example, of the Δ key. Incrementally select a portrait by operating, or decrement the portrait by operating the ▽ key). Then, step S11
In step 4, it is determined whether or not the end switch (end key) is turned on. If the end switch is not turned on, the process returns to step S110 and the same loop is repeated. And
Finally, the desired portrait corresponding to the search data is displayed on the first display unit 11. After the desired portrait corresponding to the search data is displayed on the first display unit 11, step S114
If it is detected that the end switch is turned on in step S100, the portrait selection process ends and the process returns to step S100.

B. The routine from the portrait creation process branches from step S100 to step S104, and image data is input in step S104. That is, by turning on the image input key of the input switch unit 12, the CCD camera unit 2 is set as a portrait target.
The face image data of the operator imaged in is supplied to the image data input unit 23 as a video signal. As a result, the video signal for one field supplied to the image data input unit 23 is converted from the Y / C separation unit 23a to the digital image data for one screen through the A / D conversion unit 23b. Next, in step S116, image data correction is performed. This is 1 converted to digital image data
With respect to the face image data for the screen, extra image data existing in the background portion is erased and corrected. Then, in step S118, the input image data is processed by the CPU.
21 is transferred to the input image memory 28, stored in the input image memory 28, and displayed on the second display unit 3. As a result, the face image of the operator captured by the CCD camera unit 2 is displayed on the second display unit 3. The input image displayed on the second display unit 3 is as shown in FIG. 9 (A), for example.

Next, in step S120, it is determined whether or not the face direction has been input. This is the input switch unit 12
Judgment is made based on whether or not the face direction key of is turned on. If the face direction key is not turned on, the process returns to step S100. If the face direction key is turned on, the process proceeds to step S122, and the work RA for storing the face image of the corresponding personal data.
Specify the M32 area. As a result, the face image data corresponding to the face direction at that time is stored in the work RAM 32. For example, when the face direction in the front is imaged, and when the direction is OK, the face in the front direction as shown in FIG. 9A is stored in the work RAM 32 as an input image.

Next, moving to FIG. 7, it is determined in step S124 whether or not the part key of the input switch unit 12 is turned on. If the part key is turned on, the input image corresponding to the part selected in step S126 is input. Display the cursor on the data. For example, as shown in FIG. 9B, when the part key is turned on and "nose" is designated, the cursor C is displayed on the second display unit 3 at the part "nose". On the other hand, if the region key is not turned on, the process jumps to step S126 and proceeds to step S128. In this way, each part of the face is extracted as a partial image from the captured face image, and the orientation data of the extracted partial image and the position data of each part are stored in the part pattern ROM 30 in advance as described below. By comparing and collating with, the similar portrait image to the captured image is combined and displayed by the combination for each part.

Then, in step S128, it is determined whether or not the position correction key of the input switch section 12 is turned on. This is to determine whether to correct the position of the part displayed by the cursor C on the second display unit 3.
If the position correction key is on, the display position of the cursor is moved in step S130. As a result, for example, as shown in FIG.
As shown in (B), when the body part key is turned on and the "nose" is designated, the display position of the cursor designating the "nose" moves, and the "nose" is slightly different on the screen. Move to a place. On the other hand, if the position correction key is not turned on, the process jumps to step S130 and proceeds to step S132.

In step S132, the input switch unit 12
It is determined whether or not the enlargement key of is turned on. This is the second
It is to determine whether to enlarge the size of the part displayed by the cursor C on the display unit 3. If the enlargement key is turned on, the part designated by the cursor C is enlarged and displayed in step S136. As a result, for example, as shown in FIG. 9C, when the region key is turned on and the "nose" is designated, the "nose" is enlarged at a predetermined magnification and displayed as an enlarged image K on the screen. . By enlarging and displaying the designated part in this way, it is easy to correct the enlarged part, and as a result, the characteristics of the part are displayed and the real part is displayed.
It becomes possible to create a caricature image (montage image ) that is close to the face . On the other hand, if the enlargement key is not turned on, the process jumps to step S134 and proceeds to step S136. In step S136, it is determined whether or not the matching key of the input switch section 12 is turned on. This is the second
The part in the captured image displayed by the cursor C on the display unit 3 is compared and collated with the part pattern stored in the part pattern ROM 30 in advance, and the matching frequency is ranked. By repeating the process for each part, it is possible to compositely display the portrait image that is theoretically closest to the captured image.

If the matching key is turned on, the part pattern of the portion designated by the cursor C is read from the part pattern ROM 30 in step S138. For example,
As shown in FIG. 9B, if the part key is turned on and "nose" is designated, the part pattern of the part "nose" is read from the part pattern ROM 30. Next, the number of matching dots in the image of the part pattern read in step S140 and the part designated by the cursor is counted. For example, if the part key is turned on and "nose" is specified, the image data of "nose" in the captured image read from the input image memory 28 and the part pattern ROM 30 are displayed.
The first part pattern of the "nose" read out from is matched, and the number of matching dots corresponding to the matching point of the image part pattern is counted. For example, in the example of FIG. 3, as the part pattern of “nose”, [01], [02],
... Since 50 types such as [50] are prepared in advance, the pattern designated by the pulse pattern number of [01] is read out as the first part pattern and the "nose" of the captured image Matched with image data. The number of matching matching dots between the “nose” image data in this image data and the first part pattern of the “nose” registered in advance is
It is stored in the matching frequency register in the part pattern ROM 30 corresponding to the first part pattern of "nose".

Next, in step S142, it is determined whether or not there is the next part pattern (for example, the second part pattern of "nose") to be compared. If there is a next part pattern, the process returns to step S138 to read the next (second) "nose" part pattern (second part pattern) from the part pattern ROM 30, and then similarly read in step S140. The number of matching dots in the image of the part designated by the cursor is counted with the created second part pattern. For example, if the “nose” part pattern is being collated, the second part pattern designated by the pulse pattern number [02] is read out and collated with the image data of the “nose” in the captured image. It
The matching match dot number between the “nose” image data in this image data and the second part pattern of the “nose” registered in advance is the matching frequency in the part pattern ROM 30 corresponding to the second part pattern of the “nose”. Stored in register.

Similarly, by repeating the loop of steps S138 to S142, the "nose" in the face image stored in the input image memory 28 as the portrait image object.
The image data and the matching process of each part pattern of the "nose" part registered in advance in the part pattern ROM 30 are sequentially repeated, and the number of matching match dots is stored in the matching frequency register corresponding to each part pattern. To go. When the matching process with all the "nose" part patterns is completed, it is determined in step S142 that there is no next part pattern, the determination result is NO, and the process branches to step S144. In step S144, the parts patterns are ranked according to the number of matching dots. That is, based on the number of matching matching dots with the “nose” image data stored in the matching frequency register corresponding to each part pattern, a ranking number is given from the highest matching frequency, and the corresponding ranking registers are assigned. Is stored. For example, as for the part pattern corresponding to "nose" which is one of the face parts, the number of matching matching dots and the order number are stored as shown in FIG. In this case, the number of matching match dots = [25] is the rank = [1], and the number of matching match dots = [20] is the rank = [2] ...
.., and so on, and are stored in the corresponding rank registers with the rank numbers assigned in descending order of matching frequency.

Next, in step S146, the coordinates of the designated part are stored in the corresponding part area of the work RAM 32. Next, in step S148, it is determined whether or not the above matching and ranking processing has been completed for all the parts. The above is the matching and ranking process for the part pattern corresponding to "nose" which is one of the face parts as an example. When the "nose" is finished, the process moves to the "mouth" and the "mouth" image is displayed. The matching process for all the "mouth" part patterns for the data and the ranking process of the matching frequency are executed. If the matching process and the matching frequency ranking process have not been completed for all parts, the process returns to step S124 to repeat the above loop. When the matching process and the matching frequency ranking process as described above are completed for all parts, the process proceeds to step S150. In this way, the matching process and the matching frequency ranking process are all repeated for each part, so that the part pattern is added to the image data of all parts of the face image data stored in the input image memory 28. All the part patterns stored in the ROM 30 are collated for each part, and the matching frequency is ranked.

In step S150, a portrait image , which is a montage image, is first created using the highest-ranking part pattern.
It is displayed on the display unit 11. That is, the part pattern RO
When the collation order of the face image data of each part pattern is stored in the order register of each part of M30, the respective parts (contour, hair, right eye, ...
The highest order part pattern in the mouth) is read from the part pattern ROM 30, and the image data of all parts of the face image data stored in the input image memory 28 is converted into the face image data stored in the part pattern ROM 30. On the other hand, a montage image is displayed on the first display unit 11 as a portrait image approximated at the first stage. That is, the operator operates the image input key of the input switch unit 12 to operate the CPU via the image data input unit 23.
When the face image data of the portrait person input to 21 is stored in the input image memory 28, the image data of each part (contour, hair, right eye, left eye, ... Mouth) included in the face image data is stored. The part pattern most similar to the pattern is automatically read from the part pattern ROM 31 based on the above pattern matching processing, and the first display unit 1
It is combined with 1 and displayed.

When it is desired to correct or change the portrait image while the portrait image most approximate to the face image data of the portrait person is displayed on the first display unit 11, the input switch unit is selected. Operate the 12 correction key. This process is executed after step S152. That is, it is determined whether or not the correction key of the input switch unit 12 is turned on in step S152. This is operated in the case of correcting the portrait image which is the first montage image as the next step. If the correction key is turned on, the process proceeds to step S154, while if the correction key is not turned on, a sufficient montage image is obtained at this stage.
It is determined that a portrait image has been created, step S168
Then, it is judged whether the registration key is turned on or not.

Although the above-mentioned portrait image is displayed on the first display portion 11, when the correction key is operated, the name of the part (contour, hair, right eye, left eye, ... .. (mouth) may be displayed. In that case, for example, the contour is initially set, and each time the correction key is operated, the display is changed to "hair" → "right eye" → "left eye" → ... In step S154, it is determined whether or not the part key is turned on. This is to determine the corrected portion. If the part key is not turned on, the process jumps to step S168. If the part key is turned on, the process advances to step S156 to display the cursor on the part of the part corresponding to the selected part. Thereby, the operator can confirm the designation of the corrected portion on the screen of the first display unit 11. Next, in step S158, it is determined whether or not the selection key (Δ ▽ key) of the input switch unit 12 is turned on. If the selection key is on, step S1
Proceed to step 60, and move the next-order part to the part pattern ROM3.
It is read from 0 and displayed in place of the part shown. As a result, a part pattern having a shape different from that of the past is displayed at the position of the selected portion.

Here, for example, when the pattern selection key (up key) is operated when the part "nose" to be corrected is selected and displayed by pressing the part key, a plurality of nose part patterns in the part pattern ROM 30 are selected. The second nose part pattern is designated and read out based on the part pattern matching order stored in the order register corresponding to, and is replaced with the first nose part pattern in the currently displayed portrait image to be synthesized and displayed. In this case, the part pattern designated as the above-mentioned correction target is replaced and corrected with the part pattern of the next rank every time the pattern selection key (up key) is operated, and the pattern selection key (down Each time the (△ key) is operated, the parts pattern of the previous rank is replaced and corrected. On the other hand, if the selection key is not turned on, the process jumps to step S160 and proceeds to step S162. At this time, the shape of the part pattern of the selected portion does not change.

Then, in step S162, it is determined whether or not the position correction key of the input switch section 12 is turned on. If the position correction key is turned on, the process proceeds to step S164 to move the part display position. Then, step S166.
The coordinates stored in the area of the corresponding portion of the work RAM 32 are changed with. Thereby, for example, when “nose” is selected as the part, the position of “nose” is moved and displayed on the first display unit 11. Then, step S16
Go to 8. On the other hand, if the position correction key is not turned on, the process skips steps S164 and S166 and proceeds to step S168. In step S168, it is determined whether or not the registration key of the input switch unit 12 is turned on. If the registration key is not turned on, the process returns to step S154 and the same loop is repeated. For example, the shape and position of parts other than the "nose" are corrected as necessary. When the correction is completed and the registration key is turned on, step S
Proceeding to 170, the corrected image is recorded in the work RAM 32. Then, the process returns to the first step S100.

As described above, the operator designates an arbitrary part to be corrected in the portrait image displayed on the first display section 11, and the part pattern is replaced with a desired part pattern or the position is moved. When the registration key is operated while the portrait image is corrected, the number of each part pattern forming the portrait image in the current display state is written in the work RAM 32 and registered. Therefore, thereafter, by designating the personal data, the portrait data can be read from the work RAM 32 and displayed on the first display unit 11. In that case, when the front face portrait data is read, for example, a montage image as shown in FIG.
A portrait image can be obtained. When the face orientation is changed and the portrait data on the right side is read , a portrait image as a montage image as shown in FIG. 10B is obtained. In addition, after the portrait data is registered in the work RAM 32,
When the print key is operated, the corresponding portrait part patterns corresponding to the registered portrait data are read out from the part pattern ROM 30, and the portrait image data obtained by combining these portrait part patterns is printed by the print drive unit 2.
5 is sent. Therefore, the portrait image corresponding to the portrait image data is printed on the label tape 4 by the printing unit 26.

As described above, in this embodiment, the image data of the face is picked up and read by the CCD camera unit 2, the read face image data is temporarily stored in the input image memory 28, and the face image data is stored from the stored face image data. A partial image is extracted for each part, the orientation data of the extracted partial image and the position data for each part are stored in the work RAM 32, and each part of the face is previously stored in the parts pattern ROM 30 such as hair, eyes, nose, mouth, etc. The plurality of types of part patterns are stored for each part, and the part data corresponding to the partial image and the part pattern RO are stored in the work RAM 32.
By comparing and collating a plurality of types of part patterns for each part of the face stored in M30, ranking the matching frequency, and repeating the matching process for combining each position data for each part. , CC
The portrait image which is theoretically closest to the captured image of the D camera unit 2 is displayed in a composite manner. Therefore,
Caricature corresponding to the captured image of the input face (actual face)
When creating an image , it is possible to easily create a caricature that is closer to the actual face , without the trouble of selecting each part pattern.

Next, FIGS. 11 to 16 are views showing a second embodiment of the face image forming apparatus and the face image data storage apparatus according to the present invention. FIG. 11 is a front view of an electronic notebook when the present invention is applied to the electronic notebook. In FIG. 11, the electronic notebook 51 shows a state in which the pair of main bodies 51a and 51b are opened to the left and right respectively. A CCD camera section ( face image acquisition means ) 52 is arranged above the main body 51b of the electronic notebook 51 as in the first embodiment. A second display section 53 and a data input key 54 are sequentially arranged below the CCD camera section 52 in the main body 51b of the electronic notebook 51. Second display section 53 and data input key 5
The contents of 4 are the same as in the case of the first embodiment. The second
A face image 52 captured by the CCD camera unit 2 is displayed on the display unit 53.
a is shown. Although not shown, a printing unit capable of printing a label tape is built in below the main body 51b of the electronic notebook 51 as in the case of the first embodiment.

On the other hand, on the main body 51a of the electronic notebook 51, a first display portion 61, an input switch portion 62, and an on / off key 63 are arranged from the upper side. The contents of the first display portion 61 and the on / off key 63 are the same as in the first embodiment. It should be noted that the first display unit 61 displays on the face image 61 during image processing.
a is shown. The input switch section 62 is the electronic notebook 5
This is for performing various operations required to create a portrait of a person by 1. Specifically, there are five keys of "individual", "start", "image input", "display", and "end". There is. For these keys, for example, thin sheet type key switches are used. Of the keys, the "personal", "image input", "display", and "end" functions are the first
It is similar to the embodiment. The "start" key is an operation switch for starting a moving image that moves over time. The block diagram showing the configuration of the face image creating apparatus of the second embodiment is the same as that of the above embodiment, and although not shown, C
The PU sequentially registers the moving face image in the form of the face orientation, the part number of each part of the face, the position data, etc. over time, so that the data of each registered part number, etc. can be recorded during playback. Based on this, processing is performed such that part patterns prepared in advance are combined to restore a face image. Further, the part pattern ROM stores the part pattern of each part with the contents as shown in FIG.

Next, the operation will be described. First, electronic notebook 5
1 main body 51a, 51b open to the left and right respectively on /
When the off key 63 is pressed, power is supplied to the electronic notebook 51 and the operation starts. After that, by executing the program described later, the personal data is input and the portrait creation process is performed. FIG. 12 is a flow chart showing a program for portrait creation processing using the electronic notebook 51 of this embodiment. When this program starts, first step S
At 200, it is determined whether or not the display switch is turned on. The display switch is a display key in the input switch section 62, and when the display switch is on, the first
The display unit 61 is activated to be able to display an image, and the process proceeds to step S202 to perform a process of selecting and displaying a portrait that has already been created. If the display switch is not turned on, the process branches to step S204, and a process of creating a portrait from an actual person image is executed. A. Routine of portrait selection processing First, the routine of portrait selection processing will be described. In step S202, it is determined whether search data has been input. The search data is data for selecting the created portrait and is, for example, personal data. If search data has not been input, the process returns to step S200, and if search data has been input, the process proceeds to step S206 and the corresponding data is searched from the personal data in the work RAM by the search data and displayed on the first display unit 61. indicate.

Here, FIG. 13 shows an example of the storage of the work RAM. For example, the facial features of each person corresponding to the personal data of the person are the elapsed time at the time of photographing the face (that is, the first photographing). T1 to Tn, each face image (A) corresponding to each elapsed time T1 to Tn, face image (B) to face image (N), and a part pattern for each part of the face The number and the position data (parts data) of each part are stored as parameters in predetermined areas.
Further, in this case, the part pattern numbers such as contour, hair, eyes, ... Corresponding to the elapsed times T1 to Tn at the time of imaging are, for example, [50], [02], [50] ,. Thus, the position data of parts such as contours, hairs, eyes, etc. are stored in a predetermined area, for example, [x1, y1], [x2, y2], [x
3, y3], ..., And stored in a predetermined area. [X1, y1], [x2, y2],
The values [x3, y3], ... Are coordinate values that specify the position on the screen, and are difference data from the absolute position according to the correction of the position. Therefore, when the position is not corrected, the coordinate value is [0,0].

Then, for example, by combining and synthesizing each part pattern as a face part having the part pattern number of each part of the face image (A) corresponding to the elapsed time T1 and the position data of each part, respectively. A portrait image, which is one montage image, is created. Therefore, when the input search data is personal data, the corresponding personal data is collated, and the matching personal data is selected and displayed on the first display unit 61. For example, if any of personal name: former address: ○○ telephone number: ○○ blood type: ○ hobby: ○○ is input, it is stored in the first area of FIG. 13 as the corresponding person. The personal data of the person who is instep is selected and displayed on the first display unit 61.

Then, in step S208, the register T in the CPU 21 which stores the elapsed time during the image pickup of the face (that is, the elapsed time after the first image pickup is started) is cleared to "0", and the flag K is set. Set to "1". The flag K is set when personal data is input, and reset when the end key is pressed. Then, the designated face image data in the area of the work RAM in which the portrait image (montage image ) corresponding to the personal data retrieved in step S210 is stored is read out and displayed on the first display unit 61. As a result, one of the portraits already created based on the personal data is picked up and displayed. For example, when the data related to the human instep as described above is input as the personal data, the elapsed time at the time of imaging corresponding to the instep, the face image, and the part pattern number for each part of the face shown in FIG. , One portrait based on the position data (parts data) of each part is displayed as an image. Therefore, although there are a plurality of portraits corresponding to the person's instep according to the elapsed time at the time of image capturing, the direction of the face, etc., one of them is selected and displayed. Next, in step S212, it is determined whether or not the end switch (end key) is turned on. If the end switch is not turned on, the process returns to step S200 and the same loop is repeated. Then, finally, a desired portrait corresponding to the search data is displayed on the first display unit 61.
When the desired portrait corresponding to the search data is displayed on the first display unit 61, the process proceeds to step S214, the flag K is reset to "0", and the process returns to step S200.

B. The routine of the portrait creation process branches from step S200 and proceeds to step S204, where image data is input. That is, by turning on the image input key of the input switch section 62, the CCD camera section 5 is selected as a portrait object.
The face image data of the operator captured in 2 is supplied to the image data input unit 23 as a video signal. As a result, the video signal for one field supplied to the image data input section is transferred from the Y / C separation section 23a to the A / D conversion section 2.
It is converted into digital image data for one screen via 3b. Next, in step S216, image data correction is performed. By this, 1 converted to digital image data
Excessive image data existing in the background portion of the face image data for the screen is deleted and corrected. Then
In step S218, the input image data is transferred from the CPU 21 to the input image memory 28, stored in the input image memory 28, and displayed on the second display section 53.
As a result, the face image of the operator captured by the CCD camera unit 52 is displayed on the second display unit 53.

Next, in step S220, it is determined whether or not the start key (start switch) of the input switch section 62 is turned on. If the start key is on, the flag P is set to "1" in step S222. The flag P is for starting the operation of capturing the imaged face image of the operator. In addition, the same step S222
Then, the register T in the CPU 21 that stores the elapsed time during image capturing is cleared to "0". Then, it progresses to step S224. On the other hand, if the start key is not turned on in step S220, step S222 is skipped and the process proceeds to step S224.

In step S224, the elapsed time T during image pickup is
Is determined to be a predetermined maximum value (max) or not. The maximum value (max) regulates the elapsed time at the time of imaging the face to a predetermined value, and is set to a value corresponding to the final elapsed time Tn after the first imaging is started, for example. If the elapsed time T at the time of imaging is not equal to the maximum value (max) in step S224, the process returns to step S100 and the same loop is repeated. As a result, the operator's face image data captured by the CCD camera unit 52 is supplied to the image data input unit 23 as a video signal at regular time intervals and converted into digital image data for one screen.
The corrected image is stored in the input image memory 28, and the image in the meantime is displayed on the second display section 53. And step S
When the elapsed time T at the time of imaging becomes equal to the maximum value (max) in 224, the process proceeds to step S226 and the flag P is set to "0".
And the process returns to step S200. This gives C
The capture of the operator's face image data captured by the CD camera unit 52 is completed.

FIG. 14 is a flow chart showing a program for the interrupt 1 processing. This program displays montage images that move over time,
It is repeated at regular intervals by interruption. When this interrupt routine starts, it is first determined in step S250 whether or not the flag K is set to "1". The flag K is set when personal data is input and is reset when the end key is pressed. Therefore, if personal data is input, K = 1 and the process proceeds to step S252 and ends. K if key is pressed
= 0 and the routine of this time is ended. Step S2
At 52, the elapsed time T of image display is incremented by [1]. Next, in step S254, the work RAM 3
The elapsed time in the area corresponding to the designated personal data in 2 is searched. For example, when data related to the person A is input as personal data, the elapsed time in the area is searched for the person A shown in FIG. Work RA
Each data as shown in FIG. 13 is stored in M32. If the data relating to the person A is input as personal data, this time, the elapsed time T1 in the area for the person A is T1.
The item of ~ Tn will be searched.

Then, in step S256, it is determined whether or not there is the same elapsed time (that is, T1 to Tn) as the elapsed time T of image display. That is, the elapsed time T of image display
Is equal to any of T1 to Tn. If the elapsed time T of image display is equal to any of T1 to Tn, the corresponding face image data is designated in step S258. For example, when the elapsed time T of image display is equal to T1, the face image data corresponding to the elapsed time T1 in the area of the person A shown in FIG. It is sent to the drive unit 25. As a result, the elapsed time T1 for the person A
The portrait corresponding to is displayed on the second display unit 63. In addition, a portrait corresponding to the elapsed time T1 of the person A can be printed if necessary. Step S258
After that, the process proceeds to step S260. Also, step S
If the elapsed time T of image display is not equal to any of T1 to Tn in 256, step S258 is skipped and the process proceeds to step S260.

In step S260, it is determined whether the elapsed time T of image display has reached a predetermined maximum value (max). The maximum value (max) regulates the elapsed time of image display to a predetermined value, and is set to a value corresponding to the final elapsed time Tn shown in FIG. 13, for example. Step S2
If the elapsed time T of image display is not equal to the maximum value (max) at 60, the routine of this time is ended, the routine is repeated waiting for the next interrupt timing. As a result, in the next routine, the face image data corresponding to the elapsed time T2 is designated, taken out by the CPU 21, and sent to the display drive unit 27 or the print drive unit 25.
Then, when the elapsed time T of image display becomes equal to the maximum value (max) in step S260, the process proceeds to step S262, T is returned to "0", the routine of this time is ended, and the next interrupt timing is waited. Therefore, for example, when the data related to [name: A] is input as personal data, the elapsed times T1 to Tn in the area are sequentially searched for the person A shown in FIG. 13, and the respective elapsed times T1 to Tn are searched.
Image data corresponding to are specified one after another and taken out,
The portraits corresponding to the human instep are sequentially displayed on the second display unit 63. Also, if necessary, the elapsed time T for the person A
It is also possible to sequentially print portraits corresponding to 1 to Tn.

FIG. 15 is a flow chart showing a program for the interrupt 2 processing. This program creates montage images that move over time,
It is repeated at regular intervals by interruption. When this interrupt routine starts, it is first determined in step S300 whether the flag P is set to "1". Since the flag P is for starting the operation of capturing the face image of the operator who has been imaged, when P = 0, this routine is ended and the next interrupt timing is awaited. On the other hand, when P = 1, the process proceeds to step S302, and the current timer time T (that is, the current measured value) is set to the work RAM 32 in correspondence with the part pattern of the face image.
Store in a predetermined area. The timer time T is to measure the elapsed time each time the interrupt routine is executed in order to match the elapsed times T1 to Tn in the area shown in FIG.

Next, in step S303, the timer time T is incremented by [1]. The incremented value [1] is, for example, a value corresponding to the repetition time of this interrupt routine. As a result, in the first routine, T = [1], and the process of creating a face image corresponding to the elapsed time T1 in the area is performed. Next, in step S304, the pointer N designating the face part is set to [1]. Next, in step S306, the Nth part in the image data displayed on the first display unit 61 is designated.
Therefore, at first, the hair contour is designated by the contour of the face corresponding to the pointer N = [1], and thereafter, the pointer N is incremented in step S320 to be described.
The eyes are designated as ... Next, in step S308, the Nth part pattern and specific coordinates (coordinates of position data) are read from the parts pattern ROM 30.

In the first routine, the part pattern number (eg, [50]) and position data (eg, [x1, y1] as shown in FIG. 13 corresponding to the pointer N = [1] are set for the contour of the face. ]) Is read. Then
In step S310, the coordinates corresponding to the specific coordinates in the designated part of the image data are detected. This allows
The coordinates of the part called the contour designated by the current pointer N = [1] is detected in the captured face image of the operator. Next, in step S312, the coordinates of the part pattern and the part of the image data are matched, and the number of matching dots is counted. As a result, in the present routine of pointer N = [1], the number of dots in the portion where the contour part pattern read from the area of FIG. 13 and the contour image of the image data match is counted.

Then, in step S314, the difference between the number of coincident dots and the coordinates is stored in the work RAM 32. As a result, in the routine of the pointer N = [1] this time, the number of coincident dots between the contour part pattern and the contour image of the image data and the difference in coordinates corresponding to the difference between each position of the contour part pattern and the image data are calculated. Work RAM3
Stored in 2. Next, in step S316, the number (NO.) Of the part pattern having a large number of matching dots is stored in the work RAM 32. As a result, in the routine of the pointer N = [1] this time, the part pattern number of the contour most similar to the contour image of the image data is stored in the work RAM 32, and the part of the portion called the contour when the portrait is created. The pattern has been selected.

Next, in step S318, it is determined whether or not the designation of all parts has been completed. If not completed, the process proceeds to step S320, the pointer N is incremented by [1], the process returns to step S306, and the same loop is repeated. Therefore, in the next loop, N = [2] is set and the part called the hairstyle is designated. In the same manner as above, the part pattern number of the hairstyle most similar to the hairstyle image is stored in the work RAM 32, and the portrait is drawn. A process of selecting a part pattern of a part called a hairstyle at the time of creating is performed. In the same manner, the part pattern number of the part most similar to the image of all parts of the face is stored in the work RAM 32 in the same manner as N = [3] eyes, ..., And a portrait is created. The process of selecting the part patterns of all the parts at that time is performed.

When the selection of the part patterns for all the parts is completed, the determination result of step S318 is YES.
Then, the interrupt routine is terminated and the next interrupt timing is waited for. Then, in the next interrupt routine, T = [2] corresponding to the part pattern of the face image in step S302, and a process of creating a face image corresponding to the elapsed time T2 in the area is performed. Thereafter, similarly, each time the interrupt routine is executed, T = [3], T = [4], ...
・ Elapsed time T2, T in the area in the order of T = [n]
A process of creating each face image corresponding to 3, T4, ... Tn is performed. As described above, in the second embodiment, a moving face image of the operator is captured, and the orientation of the face according to the passage of time,
Registration is performed in order in the form of part numbers and position data for each part of the face. Then, at the time of reproduction, a part pattern prepared in advance based on the registered data such as each part number is combined and restored as a face image. Therefore, Iwayu moving montage image is similar
A face image can be reproduced.

For example, as shown in FIG. 16 (A), a moving face image of an operator is captured as a front image, for example, as a front image at a time T1 from the start of image capturing, and the orientation of the face and the face at that time with respect to the captured image. The part number, position data, etc. for each part are obtained and registered together with the elapsed time T1. Similarly, when the elapsed time T2 has elapsed from the start of imaging, a moving face image of the operator is captured, for example, as a front image, and the orientation of the face at that time with respect to the captured image, the part number of each part of the face, position data, and the like are displayed. It is obtained and registered together with the elapsed time T2. Further, at the elapsed time T3, a moving face image of the operator is captured, for example, as a front image, and the orientation of the face at that time with respect to the captured image, the part number of each part of the face, position data, and the like are obtained. It is registered together with the elapsed time T3. When the operator turns sideways, for example, it is captured as a side image, and the face orientation at that time with respect to the captured image, the part number for each part of the face, position data, etc. are obtained in the same manner. It is registered together with the elapsed time Tx.

On the other hand, at the time of reproducing the montage image, the orientation of the registered face corresponding to the elapsed time from the start of image capturing for the first time T1 is corresponded to the work RAM 32 with the part number of each face part and the position data. Figure 16 (A)
It is restored as a front montage image as shown in. Thereafter, in the same manner, the registered face orientation corresponding to the elapsed time T2, the part number of each part of the face, the position data, etc. are read from the corresponding area of the work RAM 32 and the part patterns are combined, Figure 16 (B)
It is restored as a front montage image as shown in. Further, at the elapsed time T3, the corresponding registered face orientation, the part number of each part of the face, the position data, etc. are read from the corresponding area of the work RAM 32 and the part patterns are combined, as shown in FIG. It is restored as a front montage image as shown in C). Therefore, the moving face image of the operator is restored by combining the parts patterns prepared in advance based on the data of each registered part number and the like over time, and a so-called moving montage image can be reproduced.

[0068]

According to the invention of claim 1 , reproduction of a face image
Prior to the above, a partial image is extracted from the photographed face image for each part of the face, and the extracted partial image and a plurality of types of parts images stored in advance for each part of the face are extracted for each part. Based on the comparison result, the part image designation data corresponding to the part image of each part stored in advance and the coordinate position data of each part of the partial image are associated with each other based on the result of the comparison. It can be sequentially stored in the storage unit as time passes. On the other hand, when playing face images
For each part image finger stored in the storage unit,
Sequentially reads out the coordinate position data stored in association with the part image and the respective part image designation data corresponding to Teide over data over time, corresponding to the read out each part image specified data is located It is possible to sequentially display the portrait images, which are created by combining the parts images of the respective parts thus arranged, on the display unit. Therefore, over time
Therefore, it displays moving caricature images or moving montage images.
Can be shown .

Further, according to the invention of claim 3, claim
As in the case of the first aspect of the invention , prior to the reproduction of the face image, a partial image is extracted for each part of the face from the captured face image, and the extracted partial image and each part of the face are stored in advance. A plurality of types of part images are compared for each part, and based on the comparison result, part image designation data corresponding to the part image for each part stored in advance and each part of the partial image. It can be stored in the storage unit in sequence as time passes in association with the coordinate position data of. On the other hand, at the time of reproducing the face image, the storage unit stores the part image corresponding to each stored part image designation data and the coordinate position data stored in association with each part image designation data. Are sequentially read out according to the progress of, and the parts images are arranged at the positions corresponding to the respective read out parts image designation data, and the portrait image created by combining the parts images of the arranged parts is created. It can be sequentially displayed on the display unit. Therefore, it is possible to display a portrait image or a montage image that moves over time.

[Brief description of drawings]

FIG. 1 is a face image creation device and face image creation according to the present invention.
It is a front view of the 1st example of the electronic organizer to which the method is applied.

FIG. 2 is a block diagram showing a configuration of a face image creating apparatus of the embodiment.

FIG. 3 is a diagram showing an example of a part pattern of each part of a face stored in a part pattern ROM of the embodiment.

FIG. 4 is a diagram showing a matching match dot number and a rank number of a part pattern of the embodiment.

FIG. 5 is a flow chart showing a processing program for inputting / registering personal data according to the embodiment.

FIG. 6 is a flowchart showing a part of a program for portrait creation processing of the embodiment.

FIG. 7 is a flowchart showing a part of a program for portrait creation processing of the embodiment.

FIG. 8 is a diagram showing an example of stored contents of a work RAM of the same embodiment.

FIG. 9 is a diagram showing an example of a face image of the same embodiment.

FIG. 10 is a diagram showing an example of a face image of the same embodiment.

FIG. 11 is a face image creation device and face image creation according to the present invention.
It is a front view of the 2nd example of the electronic organizer to which the composition method is applied.

FIG. 12 is a flowchart showing a program for portrait creation processing according to the embodiment.

FIG. 13 is a diagram showing an example of stored contents of a work RAM of the same embodiment.

FIG. 14 is a flowchart showing a program for interrupt 1 processing of the same embodiment.

FIG. 15 is a flowchart showing a program for an interrupt 2 process of the same embodiment.

FIG. 16 is a diagram showing an example of a face image of the embodiment.

[Explanation of symbols]

1,51 electronic notebook 2, 52 CCD camera section (Face image acquisition means) 3,53 Second display 4 Data input key 11 First display 12 Input switch section 13 ON / OFF key 14 part keys 21 CPU 21 Key input section 23 Image data input section 24 Pattern matching section 26 Printing Department 28 Input Image Memory (Face Image Data Storage Means) 29 Basic position memory 30 parts pattern ROM (Parts storage means, second
Means of storage) 31 ROM 32 work RAM (Parts storage means, first memory hand
Step) 101Comparison means 102Means of creation

Continuation of the front page (56) References JP-A-2-68666 (JP, A) Mitsuo Shiono, Satoshi Takeda, Tomohiko Murayama, portrait drawing system from facial images, Journal of the Television Society of Japan, Japan Television Society, December 20, 1988, Vol. 42 No. 12, p1380-1386 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06T 11/60-13/00

Claims (4)

(57) [Claims] 1. A partial image is extracted from the photographed face image for each part of the face prior to reproduction of the face image, and the extracted partial image and a plurality of types stored in advance for each part of the face. Part images are compared for each part, and based on the comparison result, part image designation data corresponding to the part image for each part stored in advance and coordinate positions for each part of the partial image. Storage control means for associating data with each other and sequentially storing it in a storage unit as time passes, and a part image corresponding to each stored part image designation data stored in the storage unit at the time of reproduction of a face image. Coordinate position data stored in association with each part image designation data is sequentially read out as time passes,
The parts images are arranged at positions corresponding to the read out parts image designation data, and the portrait images created by combining the arranged part images of the respective parts are sequentially displayed on the display unit. A face image creation apparatus comprising: a display control unit. 2. The face image creation according to claim 1, wherein the face image is a face image acquired from any one of a CCD camera, a video camera, an image scanner, an electronic still camera, and another device. apparatus. 3. A partial image is extracted from the photographed face image for each part of the face prior to the reproduction of the face image, and a plurality of types are stored in advance for each of the extracted partial image and each part of the face. Part images are compared for each part, and based on the comparison result, part image designation data corresponding to the part image for each part stored in advance and coordinate positions for each part of the partial image. A storage control step of associating the data with each other and sequentially storing the data in the storage unit according to the passage of time; and a part image corresponding to each stored part image designation data stored in the storage unit when the face image is reproduced. Coordinate position data stored in association with each part image designation data is sequentially read out as time passes,
The parts images are arranged at positions corresponding to the read out parts image designation data, and the portrait images created by combining the arranged part images of the respective parts are sequentially displayed on the display unit. A face image creation method comprising: a display control step . 4. The face image creation according to claim 3, wherein the face image is a face image acquired from any one of a CCD camera, a video camera, an image scanner, an electronic still camera, and another device. Method.