JPH11169357A - Image processing apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing apparatus which can prepare image data which can be seen a little better than the image of the face of a user by subjective evaluation of the user without taking a person's help. SOLUTION: This image processing apparatus is provided with an input part 1 for inputting image signals, a storage means 2 for storing the inputted image as a digital image, a shape extracting means 110 for identifying a shape from the image data read out from the storage means and extracting the region, a model storing means 140 for outputting an appropriate model in accordance with a shape signal for expressing the kind of the shape discriminated by the shape extracting means 110, a transformation means 120 for transforming the shape region discriminated from the image data so as to bring it close to an appropriate model from the model storing means 140 by a certain ratio, a synthesizing means 130 for synthesizing the image data transformed by the transformation means 120 to the original image data and a displaying part 3 for displaying the synthesized image data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and more particularly, to an image processing apparatus which transforms image data input from a scanner, a camera or the like so as to approach a target image and outputs the image data to an output device such as a printer or a display unit. It relates to a processing device.

[0002]

2. Description of the Related Art In recent years, there is an apparatus called "print club (print club)" for taking a photograph and outputting the photograph as a seal. In this device, a low-resolution CCD camera is generally used in many cases, so that the seal has a low resolution and poor image quality. However, many girls, mainly female high school students, take pictures with this device. The reason is presumed to be that, because of the low resolution, even if acne is present on the face, the acne disappears and the face looks better than the real face.

[0003] Therefore, in order to create a photographing device that looks better than a real face photograph even when the resolution is high like a photograph, it is necessary to edit the image data of the read face with a so-called image editing tool. There is.

Although the purpose of use is different,
As described in Japanese Patent Publication No. 9276, a montage creation device that has a plurality of data for each part of the face such as eyes, nose, mouth, eyebrows, and selects, enlarges, reduces, moves and displays any of them. Are known. In this device, the eyes of the user,
Register the data for each part of the face such as nose, mouth, eyebrows,
When deformed and combined, image data can be created that looks a little better in the subjective evaluation of the user than the image of the real face of the user.

[0005]

However, in a so-called print club apparatus, it is possible to select a decorative frame or a celebrity to be combined with an image, but it is not possible to transform the actual image data of the user. Was not taken into account.

On the other hand, in a conventional example such as Japanese Patent Application Laid-Open No. 6-179276, it is necessary for a user to input commands such as selection, enlargement, reduction, and movement of face part data such as eyes, nose, mouth, and eyebrows. If the user does not have a certain knowledge of image processing, the time required for trial and error becomes longer, and the rotation rate of use of the photographing apparatus decreases.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus capable of producing image data that is slightly better viewed by a user's subjective evaluation than a real face image of the user without bothering the user. .

[0008]

To achieve the above object, the present invention provides an input means for inputting an image signal, a storage means for storing the input image as a digital image, and image data read from the storage means. Shape extracting means for identifying a shape from the image and extracting the area, model storage means for outputting an appropriate model according to a shape signal representing the type of the shape identified by the shape extracting means, and a shape area identified from the image data. A deforming means for deforming the appropriate model from the model storing means so as to approach the model at a certain rate; a synthesizing means for synthesizing the image data deformed by the deforming means with the original image data; and a display unit for displaying the synthesized image data. Is proposed.

In order to achieve the above object, the present invention also provides an identification means for identifying the shape of an object in input image data, and different model shapes corresponding to a large number of shapes identified by the identification means. The present invention proposes an image processing apparatus including a storage unit that outputs a model shape corresponding to a shape identified by an identification unit when pre-stored and performing image processing, and a deformation unit that approximates the identified shape to the model shape output from the storage unit. .

In order to achieve the above object, the present invention further provides identification means for identifying the shape of an object in input image data, and different model shapes each corresponding to a number of shapes identified by the identification means. Storage means for outputting model information and position information corresponding to the shape identified by the identification means when the position information where the shape is arranged is stored in advance and performing image processing; and an ideal shape output from the storage means for the identified shape. And an image processing apparatus comprising a deforming means for bringing the shape closer to position information where the shape is arranged.

More specifically, identification means for identifying hairstyles, face contours, eyes, mouths, nose, ears, and eyebrows from input human face image data, and a number of hairstyle shapes identified by the identification means Storage means for storing in advance a face contour model shape corresponding to the shape of the hairstyle identified by the identification means when the respective face contour model shapes corresponding to are stored in the image processing, and a face contour model shape read from the storage means. Contour deforming means for approaching the contour shape of the identified face, storage means for storing in advance the positional information of eyes, mouth, nose, ears, and eyebrows corresponding to a large number of face contours, and contour shape of the face having passed through the contour deforming means The corresponding eye, mouth, nose, ear, and eyebrows are extracted from the storage means according to the position information of the corresponding eyes, mouth, nose, ears, and eyebrows so that the eyes, mouth, nose, ears, and eyebrows identified by the contour shape of the face through the contour deformation means are brought closer to the position information. Synthetic hands to place and combine When an image processing apparatus comprising a output means for outputting the synthesized image data.

Identification means for identifying hairstyles, face contours, eyes, mouth, nose, ears, and eyebrows from input human face image data;
A storage unit that outputs a face contour model shape corresponding to the shape of the hairstyle identified by the identification unit when preliminarily storing and performing image processing on each face outline model shape corresponding to a number of hairstyle shapes identified by the identification unit; Contour deforming means for bringing the identified face contour shape closer to the face contour model shape read from the storage means, and eyes, mouths, and mouths corresponding to a large number of face contours, respectively.
The storage means for storing in advance the shape and position information of the nose, ears and eyebrows, and the shape and position information of the eyes, mouth, nose, ears and eyebrows corresponding to the contour shape of the face having passed through the contour deformation means. Eyes, mouth, nose, ears,
Deformation means for deforming the eyebrows so as to approximate the shapes of the eyes, mouth, nose, ears, and eyebrows of the memory means, and eyes, mouths,
Image processing comprising a synthesizing means for arranging the nose, ears and eyebrows so as to approach the position information from the storage means and synthesizing them together with the contour shape of the face having passed through the contour deforming means, and an output means for outputting the synthesized image data It can also be a device.

In any of the image processing apparatuses, there is provided a means for manually designating an area when the area matching the model shape cannot be detected by the identifying means or the shape extracting means or when the intention of the user to select the area is reflected. be able to.

Further, it is possible to provide a means for manually setting the rate of deformation so as to approach the model shape.

In the present invention, when creating image data that looks a little subjectively better than a photograph of the real face of the user, the face images of a large number of specific persons that the user considers subjectively good Data is input in advance from the input unit, and is separated into parts such as hairstyle, face outline, eyes, nose, mouth, and eyebrows by a shape region extraction means, and a shape signal representing the type of the identified shape is set as an address, and is input via a selector. Then, each part is stored as a model in the model storage means.

The user's face image data, which is the image to be synthesized, is processed in the same manner as when the model is stored until the shape region is extracted. However, at the time of synthesis, the model storage means performs only read processing, and is fixed so that data flows from the model storage means to the transformation means. Therefore, the model data stored in the model storage mode is read out from the model storage means for the address of the shape signal for each part from the shape area extraction means. The model data read from the model storage means is sent to the deformation means.

The deformation means deforms the model data from the model storage means so that the shape area extracted by the shape area extraction means approaches a certain ratio.

The synthesizing means synthesizes the image data deformed by the deforming means with the original image data, displays the synthesized image on the display unit, or outputs the synthesized image to the printer.

As a result, the shape of the face which the user subjectively considers to be good is approached, so that image data which can be seen better than the image of the real face of the user can be created without bothering the human hands. .

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, an embodiment of an image processing apparatus according to the present invention will be described with reference to FIGS.

Embodiment 1 FIG. 1 is a block diagram showing a system configuration of Embodiment 1 of an image processing apparatus according to the present invention.
The image processing apparatus includes an input unit for inputting an image signal, a storage unit for storing an input image as a digital image, and a shape extraction for identifying a shape from image data read from the storage unit and extracting a region thereof. Means 110 and a model storage means 140 for outputting an appropriate model in accordance with a shape signal representing the type of shape identified by the shape extraction means 110
And a transforming unit 120 for transforming the shape area identified from the image data so as to approach a suitable model from the model storing unit 140 at a certain ratio, and combining the image data transformed by the transforming unit 120 with the original image data. Means 130
And a display unit 3 for displaying the combined image data. This image processing apparatus has two modes: a model storage mode and a synthesis mode.

First, the flow of signals in the model storage mode will be described. For example, to create image data that looks a little better subjectively than a photograph of the user's real face,
The input unit 1 captures, from a scanner, a camera, or the like, face image data of a specific large number of people who are subjectively good for the user, amplifies the data with an amplifier, converts the data into a digital image with an AD converter, and To be stored. Shape region extraction means 11
Reference numeral 0 designates a model storage unit 1 which is separated into parts such as a hairstyle, a face contour, eyes, a nose, a mouth, and eyebrows, and uses a shape signal b, that is, a signal representing the type of the identified shape, as an address for each part.
40. In the case of the model storage mode, the selector is fixed so that the shape area signal a is stored in the model storage means 140, and the model storage means 140 performs only a storage process.

Next, the flow of signals in the combining mode will be described. The face image data of the user, which is the image to be combined, is processed up to the shape region extracting means 110 in the same manner as in the model storage mode. However, the model storage unit 140 performs only the reading process in the synthesis mode, and the selector is fixed so that data flows from the model storage unit 140 to the transformation unit 120. Therefore, the model data stored in the model storage mode is read from the model storage means 140 with respect to the address of the shape signal b for each part from the shape area extraction means 110. The model data read from the model storage means 140 is stored in the selector
20.

The transforming means 120 includes a model storing means 140
Is deformed so that the shape area extracted by the shape area extraction means 110 approaches the model data from.

The synthesizing unit 130 synthesizes the image data deformed by the deforming unit 120 with the original image data, and displays the synthesized image on the display unit 3. Further, the composite image can be output to a printer and printed out on paper.

FIG. 2 is a block diagram showing the configuration of an embodiment of the shape region extracting means 110 of the image processing apparatus of FIG. 1, and FIG. 3 is a diagram showing an example of black data connection information in the present invention. is there. The image data stored in the storage unit 2 enters the edge detection unit 111. Here, in order to extract the edges of the image, the edges are binarized into black and the non-edges are binarized into white. This binarized data is sent to the shape obtaining unit 112.
In an N × M pixel matrix. The shape pattern storage means 113 stores hairstyles, face outlines, eyes, nose,
The shapes of the mouth, eyebrows, and the like are stored in advance as black data connection information as shown in FIG.
The link information of the black data is given to the shape obtaining means 112 in the order of eyes, nose, mouth, and eyebrows.

The shape obtaining means 112 performs pattern matching between the obtained black data connection information of the N × M pixel matrix of the binary data and the black data connection information provided from the shape pattern storage means 113. I do. First,
The link information of the given hairstyle black data is obtained by the shape obtaining means 1
12, the area is extracted by the area extracting means 114, and the hair shape area signal a
To the deforming means 120. At this time, a shape signal b indicating that the detected area is a hairstyle is also sent to the model storage means 140 at the same time.

Thereafter, when the area extracted by the area extracting means 114 has been sent, the address c of the shape pattern storing means 113 is updated, and the link information of the black data of the face outline is given to the shape obtaining means 112. The rest continues in the same way.

Then, the address is updated, and the eyes, nose, mouth,
Area data of face parts in the order of eyebrows (shape area signal a)
Is given to the deformation means 120, and the identification signal (shape signal b) is given to the model storage means 140.

FIG. 4 is a block diagram showing a configuration of one embodiment of the edge detecting means 111 in the image processing apparatus of FIG. The edge detection unit 111 performs a 3 × 3 pixel matrix 1 for edge detection on the input image data.
With the central pixel A as a processing pixel in the pixel region of 111, the average value 1113 of peripheral pixels (addition value of peripheral pixels 1112/8)
The difference 1114 is obtained between the pixel value and the processing pixel A, and the absolute value of the difference value is obtained by the absolute value calculation unit 1115. If the absolute value of the difference is larger than a predetermined threshold, the processing pixel A
Is black as an edge, otherwise white as a non-edge and edges are detected.

FIG. 5 is a block diagram showing the configuration of an embodiment of the shape obtaining means 112 in the image processing apparatus of FIG. The binarized data provided from the shape pattern storage unit 113 is stored in the pixel matrix 1122. Further, the binary data from the edge detection unit 111 is stored in the pixel matrix 1121. The comparator 1123 performs pattern matching between the patterns of the pixel matrix 1121 and the pattern of the pixel matrix 1122. When the pattern matching is performed, the shape signal b is enabled and output, and the image data and the binarized data at that time are extracted by the area extracting means. 114
And output. After sending the binary data, the address counter for generating the address c is incremented,
Update. The AND circuit 1124 generates the shape signal b by the logical product of the value of the address counter and the enable signal by matching.

The area extracting means 114 shown in FIG. 2 can be realized by a selector. The image data and the binarized data output from the region extracting means 114 are valid only when the shape signal b is enabled.

FIG. 6 is a block diagram showing the configuration of an embodiment of the deformation means 120 in the image processing apparatus of FIG.
FIG. 7 is a diagram showing an example of how an actual eye image is deformed using a model by the deforming means of FIG. The deforming means 120 first calculates the width of the area, that is, the maximum value in the vertical direction and the maximum value in the horizontal direction, based on the binarized data from the shape area extracting means 110. Based on the calculation result, the vertical and horizontal widths of the model image data from the model storage unit 140 are enlarged or reduced. The method of enlarging or reducing may be executed by simply increasing or removing the same pixel every several pixels, or may be executed by a projection method.

First, the process is started from the binarized data from the shape region extracting means 110. The binarized data and the model image data are placed at the coordinates of the matrix, and the difference between the binarized data and the model image data is determined for each pixel. This difference result is multiplied by a certain ratio, for example, 10%, and the result obtained by multiplying the binarized data is added and deformed. Next, the image data from the extracting means 110 is enlarged or reduced for each row and each column so as to be deformed so as to have the shape of the binarized data in the shape of the deformed area. The ratio by which the difference result is multiplied may be changed to a value of, for example, 5% or 15% depending on parts such as eyes, nose, and mouth.

The combining means 130 combines the transformed image data with the original image data. The synthesizing means 130
It can be realized with memory. If the ratio of multiplying the difference result is large, the deformation will be large, and the face will not resemble the real face of the user, and the user will not feel the actual face of the user, so it will be suppressed to a slight deformation There is a need.

Embodiment 2 FIG. 8 is a block diagram showing the system configuration of Embodiment 2 of the image processing apparatus according to the present invention.
In the image processing apparatus shown in FIG. 1, when a face part cannot be detected by the shape area extracting unit 110, the synthesized image is incomplete. Therefore, if the face part is not properly detected by the shape area extracting means 110, the shape area extracting means 110
The image area and the face part may be selectable.

Also, the rate of deformation by the deformation means can be specified by a command input from the user when there is a particular user.

In any of the embodiments, when the input image data is a person, the image data is read out from the storage means rather than being deformed and synthesized first with particular attention to details such as eyes, mouth, nose, ears, and eyebrows. First, the contour shape of the identified face is brought closer to the face contour model shape, and the position information of each corresponding part is extracted from the storage means according to the deformed contour shape,
It is more natural to arrange and combine the identified eyes, mouth, nose, ears, and eyebrows in accordance with that.

At this time, not only the positions of the eyes, mouth, nose, ears, eyebrows and the like are changed, but also the image data of each part may be deformed.

In the above example, the target of image processing is an image of a human face. However, for image data of an object other than a human face, target image data is stored in the model storage unit 14.
If stored in 0, image data that is closer to target image data than real image data can be obtained.

[0041]

According to the present invention, there is provided an image processing apparatus which can create image data which can be seen a little better by a subjective evaluation of a user than an image of a real face of the user without bothering humans. Is obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a system configuration of an image processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration of an embodiment of a shape region extracting unit of the image processing apparatus of FIG. 1;

FIG. 3 is a diagram showing an example of black data connection information according to the present invention.

FIG. 4 is a block diagram illustrating a configuration of an embodiment of an edge detecting unit in the image processing apparatus of FIG. 1;

FIG. 5 is a block diagram illustrating a configuration of an embodiment of a shape obtaining unit in the image processing apparatus of FIG. 1;

FIG. 6 is a block diagram illustrating a configuration of an embodiment of a deformation unit in the image processing apparatus of FIG. 1;

FIG. 7 is a diagram showing an example of how an actual eye image is deformed using a model by the deforming means of FIG. 6;

FIG. 8 is a block diagram illustrating a system configuration of an image processing apparatus according to a second embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input part 2 Storage means 3 Display part 4 Printer 100 Image processing part 110 Shape area extraction means 120 Deformation means 130 Synthesis means 140 Model storage means 111 Edge detection means 112 Shape acquisition means 113 Shape pattern storage means 114 Area extraction means 121 Enlargement, Reduction processing means 122 Subtractor 123 Multiplier 124 Adder 125 Area size detection means 126 Area size detection means 127 Enlargement / reduction processing means 1111 Edge detection matrix 1112 Adder 1113 Average value calculation means 1114 Subtractor 1115 Absolute value calculation means 1116 Threshold value 1117 Comparator 1121 Pixel matrix 1122 Pixel matrix 1123 Comparator 1124 AND circuit 1125 Address counter a Shape area signal b Shape signal c Address of shape pattern storage means Scan

Claims (7)

[Claims] An input unit for inputting an image signal; a storage unit for storing the input image as a digital image; a shape extraction unit for identifying a shape from image data read from the storage unit and extracting a region thereof; A model storage unit that outputs an appropriate model in accordance with a shape signal representing the type of the shape identified by the shape extraction unit; and a shape region identified from the image data is brought closer to the appropriate model from the model storage unit at a certain ratio. An image processing apparatus comprising: a deforming means for deforming the image data as described above; a synthesizing means for synthesizing the image data deformed by the deforming means with the original image data; 2. An identification means for identifying the shape of an object in input image data, and said identification means for preliminarily storing different model shapes corresponding to a large number of shapes identified by said identification means and performing image processing. An image processing apparatus, comprising: a storage unit that outputs a model shape corresponding to the shape identified in (1), and a deformation unit that approximates the identified shape to the model shape output from the storage unit. 3. An identification means for identifying a shape of an object in input image data, and different model shapes corresponding to a large number of shapes identified by the identification means and position information where the shapes are arranged are stored in advance. Storage means for outputting a model shape and position information corresponding to the shape identified by the identification means at the time of image processing; and an ideal shape output from the storage means for the identified shape and a position where the shape is arranged. An image processing apparatus comprising a deforming unit that approaches information. 4. A hairstyle,
Identification means for identifying the outline of the face, eyes, mouth, nose, ears, and eyebrows; and a method for storing the respective face outline model shapes corresponding to a number of hairstyle shapes identified by the identification means in advance and performing image processing. Storage means for outputting a face contour model shape corresponding to the shape of the hairstyle identified by the identification means; contour deformation means for bringing the identified face contour shape closer to the face contour model shape read from the storage means; Storage means for storing in advance the position information of the eyes, mouth, nose, ears and eyebrows corresponding to the contour; and eyes, mouth, nose, and nose corresponding to the contour shape of the face having passed through the contour deformation means. Synthesizing means for extracting positional information of ears and eyebrows, arranging and identifying the identified eyes, mouth, nose, ears, and eyebrows in the contour shape of the face having passed through the contour deforming means so as to approach the positional information; data The image processing apparatus comprising a output means for outputting. 5. A hairstyle and a hairstyle from input human face image data.
Identification means for identifying the outline of the face, eyes, mouth, nose, ears, and eyebrows; and a method for storing the respective face outline model shapes corresponding to a number of hairstyle shapes identified by the identification means in advance and performing image processing. Storage means for outputting a face contour model shape corresponding to the shape of the hairstyle identified by the identification means; contour deformation means for bringing the identified face contour shape closer to the face contour model shape read from the storage means; Storage means for storing in advance the shape and position information of the eyes, mouth, nose, ears, and eyebrows corresponding to the contour; and eyes, mouth, and mouth corresponding to the contour shape of the face having passed through the contour deformation means. The eye, mouth, nose,
Deformation means for deforming the ears and eyebrows so as to approximate the shapes of the eyes, mouth, nose, ears, and eyebrows of the storage means; and disposing the eyes, mouth, nose, ears, and eyebrows through the deformation means from the storage means. An image processing apparatus comprising: synthesizing means for arranging the image so as to be close to the position information and synthesizing the image with the contour shape of the face having passed through the outline deforming means; and outputting means for outputting synthesized image data. 6. The image processing apparatus according to claim 1, wherein the identification unit or the shape extraction unit cannot detect a region that matches the model shape, or an intention of the user to select a region. An image processing apparatus comprising means for manually designating an area when the image is reflected. 7. The image processing apparatus according to claim 1, further comprising means for manually setting a rate of deformation so as to approach a model shape.