JPH0896111A - Device and method for hair style composition - Google Patents

Abstract

PURPOSE: To easily composite a hair style with a natural face by selecting desired face component data out of a face component data base, editing the sizes and positions of images of selected hair style data and face component data, and superposing the face image and one of a deformed hair style image and a face component image one over the other. CONSTITUTION: A face area specification part 2 extracts a necessary face area from human body image data inputted from an input part 1 to obtain face area data. A hair style selection part 5, on the other hand, selects desired hair style data and face area data out of a hair style data base 3 and the face component data base 4. An image editing part 7 edits the selected hair style data and face component data to size matching the face area data. A composition part 8 as a superposition part displays the face area data, hair style data, and face component data one over another. Consequently, a new hair style can easily be composited above a face without an incomplete hair style composite display having the projection of the original hair style or a missing face component.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a hairstyle synthesizing apparatus and a hairstyle synthesizing method, and in particular, when deciding a customer's hairstyle at a beauty salon or the like, various hairstyles are synthesized and displayed on the customer's face in advance. The present invention relates to a hairstyle synthesizing apparatus and method for evaluating a finished state.

[0002]

2. Description of the Related Art As a hairstyle synthesizing apparatus described above, Japanese Patent Laid-Open No.
1-65576. This is a first recording / reproducing device for recording / reproducing a hairstyle or a beard shape, a television receiver for displaying a reproduction signal of the first recording / reproducing device, and an image of the television receiver. A first video camera, a second video camera for picking up an image of a person, a second recording / reproducing device for recording / reproducing the person increase from the second video camera, and a second
And an image synthesizing circuit for synthesizing the output signal of the recording / reproducing apparatus.

Thus, the size and position of the hairstyle to be combined can be adjusted in the still image of the person obtained from the video camera,
A composite image can be created by superimposing an image of a hairstyle, spectacles, etc., on an image of a person.

[0004]

By the way, in the above-described conventional hairstyle synthesizing apparatus, when the original hair of a person is long and the synthesized hair is short, as shown in FIG. There is a problem that the original hair 32 protrudes from under the hair 31 combined with the person 30.

If a part of the face such as ears is hidden by the original hair of the person and the ears etc. appear in the synthesized hair, as shown in FIG. There is a problem that a part of the face 33 such as the ears 35 is not displayed in the combined image, resulting in an unnatural image.

Therefore, when actually used in a beauty salon or the like, for example, when photographing a person, it is necessary to gather hair so that the face is not hidden and all the face parts are photographed. Conceivable.

However, in such a method, there is a problem that it takes a long time to photograph and a burden is placed on the customer. Further, in the above-described conventional hairstyle synthesizing device, the facial image can be input only by the video camera, and for example, the facial image cannot be input from the portrait image, so that the application range is limited. Will be things.

[0008] Therefore, an object of the present invention is to provide a hairstyle synthesizing device and a hairstyle synthesizing method that can easily synthesize a hairstyle on a natural face without the original hair protruding and there is no hidden hidden portion. And

[0009]

A hairstyle synthesizing apparatus according to a first aspect of the present invention includes an input section for inputting a person image, a face area extracting section for extracting a face area from the input image, and a plurality of hairstyle images. A hairstyle database that stores data, a face part database that stores image data of parts that make up a face, a hairstyle selection unit that selects desired hairstyle data from the hairstyle database, and a desired face part from the face part database A facial part selection section, an image editing section for editing the size and position of the selected hairstyle data and facial part data image,
And a superimposing unit that superimposes the face image and at least one of the deformed hairstyle image and the face part image.

A hairstyle synthesizing apparatus according to a second aspect of the present invention is
An input unit for inputting a person image, a face area extraction unit for extracting a face image from the input image, a hairstyle database storing data of a plurality of hairstyles, a face component database storing data of components constituting a face, Select the desired hairstyle data from the hairstyle database, the hairstyle selection part to select the desired facial part data from the facial parts database, and the size and position of the selected hairstyle data and facial part data image. An image editing unit for editing, a superimposing unit for superimposing the face image and at least one of the deformed hairstyle image and face part image, and wearing including eyeglasses, beards, accessories, and wearing item image data. Further superimposes the product database and the mounted product image selected from the mounted product image database on the image obtained by the superposing unit. And means for displaying images.

A hairstyle synthesizing apparatus according to a third aspect of the present invention is
The image editing apparatus provided in the hairstyle synthesizing apparatus according to the first and second aspects further includes a color conversion unit that converts the color of the hair selected by the hairstyle selection unit.

A fourth aspect of the present invention is a method for editing and superimposing hairstyle image data for a desired hairstyle on the person image data input from the input unit,
Obtain the face area data by extracting the necessary face area from the person image data, select the desired hairstyle data from the hairstyle database, select the face part data missing from the face area data from the face part database, Edit the data and face part data to a size that matches the above face area data,
A hairstyle synthesizing method characterized by displaying the face area data, the hairstyle data, and the face part data in an overlapping manner.

[0013]

According to the present invention, the face area data is obtained by extracting the necessary face area from the person image data input from the input section, and the hairstyle selecting section extracts the desired hairstyle data and face part data from the hairstyle database and the face part database. Then, the image editing unit edits the selected hairstyle data and face part data to a size that matches the face area data. Then, the superposing section superimposes and displays the face area data, the hairstyle data, and the face part data. As a result, instead of the incomplete hairstyle synthesis display in which the original hairstyle protrudes or the facial parts are missing, a complete hairstyle synthesis display having a new hairstyle having all the face configurations to be displayed can be performed.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a hairstyle synthesizing device according to the present invention will be described in detail below with reference to the drawings. 1 to 9 show an embodiment of a hairstyle synthesizing apparatus according to the present invention. FIG. 1 is a block diagram showing the configuration of the hairstyle synthesizing apparatus according to this embodiment. In this embodiment, the hairstyle synthesizing apparatus includes an input unit 1 for inputting a person image, a face area extraction unit 2 for extracting a face area from the input image, and a hairstyle storing image data of a plurality of hairstyles as shown in FIG. A database 3, a face part database 4 that stores image data of parts forming a face, a hairstyle selection unit 5 that selects desired hairstyle data from the hairstyle database 3, and a desired face part is selected from the face part database 4. The facial part selecting unit 6 includes an image editing unit 7 that edits the size and position of the selected image data, and a combining unit 8 that is a superposing unit that superimposes the face image and the deformed hairstyle image. .

Further, the hairstyle synthesizing apparatus according to the present embodiment outputs an synthesizing result, a command designating section 10 for designating a work area command, an image memory 11 for temporarily storing an image, a control program and the like. Memory 12 for storing
It has a frame memory 13, a D / A converter 14, a controller 15 such as a microcomputer for controlling the entire apparatus, and a bus 16 for connecting each unit.

The input unit 1 is composed of a color video camera 21 for photographing a customer as an object, an A / D converter 22 and a scanner 23 for reading a person image from a photograph or the like.
The three digital color signals of are output via the frame memory 13.

The face area designating section 2 is for designating an area unnecessary for hairstyle synthesis in a human image. In the present embodiment, it is not necessary to specify a strict area, and it is possible to roughly specify the area by a rectangle or an ellipse to the extent that the original hair does not protrude during synthesis.

The hairstyle database 3 stores the shapes of various hairstyles. The face parts database 4 is
In addition to the main parts of the face such as the eyes, nose, mouth, ears, and eyebrows, the outline of the cheek, the chin, the neck, the forehead, etc. are stored so that various situations can be dealt with. In addition, images of wearing items such as eyeglasses, beards, and accessories may be provided as needed.

The image editing unit 7 has an affine transformation circuit and is for performing graphic transformation such as image deletion, enlargement / reduction, rotation, reversal, movement and color conversion. . The output unit 9 is
It is composed of a color monitor 24 for displaying the composition result and a color printer 25 for outputting a hard copy of the composition result.

Further, the command control unit 10 uses the mouse 2
6, having a tablet 27 and a keyboard 28. Next, the operation of the hairstyle synthesizer according to this embodiment will be described.

FIG. 2 is a flow chart showing the operating state of the hairstyle synthesizing apparatus according to this embodiment. First, the target person image is taken in (S1). The input of this person image is
The person is photographed by the video camera 21 and the still image of this video image is used, or the photograph of the person is taken in by the scanner 23. The image taken in is stored in the image memory 12 as a digital signal of R, G and B colors.
Stored in.

Then, it is judged whether it is necessary to remove extra hair from the stored image (S2). This decision is
When synthesizing a new hairstyle on a person image, this is performed when it is likely that there is hair protruding from the newly overlaid hairstyle on the person image.

When it is necessary, the person image is displayed on the monitor 24.
The face area is roughly specified by using the mouse 26 of the command specifying unit 10 and the image outside the specified area is deleted by the face area specifying unit 2 (S3). The image is deleted by using the image deleting function of the image editing unit 7.

Next, a hairstyle is selected (S4). The hairstyle image stored in the hairstyle database 3 is displayed on the monitor 24, and the desired hairstyle is selected by the hairstyle selection unit 5 using the pointing device such as the mouse 27 of the command designation unit 10 according to the customer's designation. To do.

At this time, for hairstyles that are not laterally symmetrical such as horizontal division, left-right inversion is performed by the left-right inversion function of the image editing unit 7, and hair color can be changed by the color conversion function of the image editing unit 7. .

Further, it is judged whether or not a face part needs to be selected (S5). In the input image, ears, eyebrows,
If face parts such as cheeks and forehead are hidden by hair, it is necessary, and the face parts are selected (S6), and the size and position of the face parts are adjusted.

Specifically, the face parts stored in the face part database 4 are displayed on the monitor 24, and a pointing device such as the mouse 27 of the command designating section 10 is used. Select. In this example,
You can also choose accessories.

The person image after the face area extraction and the selected face part image are displayed on the monitor 24, and the image editing unit 7
By using the moving function and the enlargement / reduction function of (1), the face parts are deformed and arranged so as to be optimal for the face (S7).

Further, the hair size and position are adjusted (S8). The person image and the hair image are displayed on the monitor 24, the moving function of the image editing unit 7 is used to move the combined hairstyle so as to overlap the head of the person image, and the scaling function is used to move the hair region. Is scaled so that it is optimal for the face, and is optimized.

Finally, the synthesizing unit 8 creates a synthetic image by superimposing the hair image, the face part image, and the person image (S9) and displays them on the monitor 24. Make a copy.

Next, a specific example of performing hairstyle synthesis using the hairstyle synthesis device according to this embodiment will be described. 3 to 6 show a first example of hairstyle synthesis.

In this example, the long hair 42 as shown in FIG. 4 is combined with the person image 40 of the short hair 41 as shown in FIG. In this case, the long hair 42 and the short hair 4 of the person
Since 1 is hidden, the face area designation processing (S3) is not necessary.
Since it is not necessary to add a face part, face part selection (S
There is no need for 6) and (S7).

First, a person is photographed and as shown in FIG.
The person image 40 is captured (S1). Next, the hairstyle is selected, and the long hair image 42 as shown in FIG. 4 is selected and captured (S
4). When the long hair image 42 is overlaid on the person image 40,
As shown in FIG. 5, the long hair 42 may be large.
Therefore, the portion of the long hair 42 is reduced and moved (S8).
Then, as shown in FIG. 6, the size and position of the long hair 42 are optimized, and a person image 40 of the long hair 42 is obtained. The processing is ended by synthesizing these, displaying and printing.

7 and 8 show a second example of hairstyle synthesis. In this example, contrary to the first example, the short hair 44 as shown in FIG. 3 is combined with the long hair person shown in FIG.

In this case, if they are combined as they are, in the conventional example, the long hair protrudes from the short hair as shown in FIG. 10, and the ears are not displayed on the human figure as shown in FIG. Therefore, face area designation processing (S3) and face part selection (S6) and (S7) are performed.

First, a person is photographed and a person image is captured (S1). Next, as shown in FIG. 8, the person image 45 is displayed on the monitor 24, the mouse 26 of the command designating section 10 is used to roughly designate the face area 46, and the face area designating section 2 designates the outside of the designated area. The image is deleted (S3).

Then, the hairstyle is selected and the short hair 44 as shown in FIG. 7 is selected and taken in (S4). Since the ears are hidden by the hair, the ears 47 are selected as face parts (S
6) Attach ears 47 to face area 46 as shown in FIG.

Then, the hairstyle is superimposed on the person image, and the hair portion is moved while being reduced (S8) to optimize the size and position of the hair portion to obtain a short-haired person image. The processing is ended by synthesizing these, displaying and printing.

Therefore, according to the present embodiment, when the hairstyle of the target person is long hair and a short hair image is synthesized,
Since the protruding hair part is deleted in advance and synthesized, it is possible to freely synthesize a selected hairstyle without limitation to the hairstyle. In addition, since face parts hidden in the input image are supplemented and combined, a natural combined image can be obtained.

[0040]

As described above, according to the present invention,
The face area data extracted from the necessary face area, the selected and edited hairstyle data, and the face part data selected and edited due to lack of the face area data are combined, so the original hairstyle may not protrude. Instead of the incomplete hairstyle synthesis display in which the face parts are missing, a natural hairstyle synthesis display having a new hairstyle having all the face configurations to be displayed can be performed.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a hairstyle synthesis device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing the operation of the hairstyle synthesis device shown in FIG.

FIG. 3 is a diagram showing an input image of an example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1.

FIG. 4 is a diagram showing an input image from a hairstyle database of an example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1.

5 is a diagram showing a display image before a deformation process of an example of hairstyle synthesis by the hairstyle synthesis device shown in FIG.

FIG. 6 is a diagram showing a display image after completion of an example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1;

FIG. 7 is a diagram showing an input image from a hairstyle database of another example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1.

8 is a diagram showing a state of face area designation processing in another example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1;

9 is a diagram showing a state of a face part deformation process of another example of hairstyle synthesis by the hairstyle synthesis device shown in FIG. 1. FIG.

FIG. 10 is a hairstyle synthesis diagram showing the problems of the conventional example.

FIG. 11 is a hairstyle synthesis diagram showing another problem of the conventional example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input part, 2 ... Face area designation part, 3 ... Hairstyle database, 4 ... Face part database, 5 ... Hairstyle selection part, 6 ... Face part selection part, 7 ... Image editing part, 8 ... Compositing part (superposition part) ).

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06F 15/62 380 15/66 450

Claims (4)

[Claims] 1. A hairstyle synthesizing device for synthesizing a desired hairstyle on a face image of a person, an input section for inputting a person image, a face area extracting section for extracting a face image from the input image, and a plurality of hairstyle data stored therein. Hairstyle database, face parts database that stores face parts data that make up a face, hairstyle selection unit that selects desired hairstyle data from the hairstyle database, and face parts selection that selects desired face parts data from the face parts database Section, an image editing section for editing the size and position of images of the selected hairstyle data and face part data, and superimposing the face image with at least one of the deformed hairstyle image and face part image A mating section,
A hairstyle synthesizing device comprising: 2. A hairstyle synthesizing device for synthesizing a desired hairstyle on a face image of a person, wherein an input section for inputting a person image, a face area extracting section for extracting a face image from the input image, and data for a plurality of hairstyles are provided. A stored hairstyle database, a face part database that stores the data of the parts that make up the face, a hairstyle selection unit that selects the desired hairstyle data from the hairstyle database, and a face part that selects the desired face part data from the face part database A selection unit, an image editing unit that edits the size and position of the images of the selected hairstyle data and face part data, and the face image and at least one of the deformed hairstyle image and face part image are superimposed. An overlaying section, a wearables database storing wearables image data including glasses, beards, and accessories; and an image obtained by the overlaying section. Hairstyle synthesizing apparatus characterized by comprising a means for displaying the image further superimposed selected attached product image from said mounting product image database for. 3. The hairstyle synthesizing apparatus according to claim 1, wherein the image editing unit includes a color conversion unit that converts the color of the hair selected by the hairstyle selection unit. 4. The person image data input from the input unit,
In a method of editing and displaying hairstyle image data for a desired hairstyle in a superimposed manner, a necessary face area is extracted from the person image data to obtain face area data, and desired hairstyle data is selected from the hairstyle database, Face part data lacking in face area data is selected from the face part database, the hairstyle data and the face part data are edited to a size that matches the face area data, the face area data, the hairstyle data, A method for synthesizing a hairstyle, characterized by displaying the face part data in an overlapping manner.