JP2006277354A - Face image operation device and face image operation program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a face image operation device and a face image operation program for changing the age impression of a face image by operating the texture of skin. <P>SOLUTION: The data of a face image are stored in an external storage device (step S1). Whether or not an aging operation has been designated by an operator is judged (step S2). When the aging operation is designated, the amplitude of the gradation waveform of the face image stored in the external storage device is enlarged (step S3). Then, a fine change is added at random for every dot of the gradation waveform (step S4). When the juvenescent operation is designated in the step S2, the amplitude of the gradation waveform of the face image stored in the external storage device is reduced (step S5). The gradation level is changed at random within the extent of amplitude for every dot of the gradation waveform (step S6). Finally, the face image after operation is displayed at a display device, and printed by a printer (step S7). <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a face image operation device and a face image operation program.

Conventionally, various apparatuses and methods for synthesizing face images have been proposed (see, for example, Patent Document 1). For example, face images with different age impressions are synthesized by manipulating the shape of facial parts (facial parts) such as eyes, nose and mouth.
Japanese Patent No. 3302122

  When a person looks at another person's face, various impressions such as age, facial expression, and health status can be obtained based on various information obtained from the face as well as the facial parts. The information obtained from the face includes various information such as the shape of the entire face, the shape of the face parts, the hairstyle, and the texture of the skin. When a person estimates the age or the like from the face of another person, it is considered that the information is integrated.

  In particular, since skin occupies most of the face area, skin texture is important in recognizing the age impression of the face. In the image processing for manipulating the face age impression, it is considered that the skin texture manipulation is indispensable for the natural age impression manipulation.

  However, the conventional image processing technique for manipulating the age impression of the face cannot manipulate the texture of the skin.

  An object of the present invention is to provide a face image operation device and a face image operation program that can change the age impression of a face image by operating the texture of the skin.

(1)
A face image manipulation device according to a first aspect is a face image manipulation device for manipulating a face image, the storage means storing a human face image as data representing a gradation level, and the data stored in the storage means And operating means for generating an age-operated face image by operating at least one of the amplitude and frequency of the waveform representing the change in the gradation level of the face image.

  In the face image operation device according to the present invention, the human face image is stored as data representing the gradation level in the storage means. Based on the stored data, at least one of the amplitude and the frequency of the waveform representing the change in the gradation level of the face image is operated by the operating means. Thereby, an age-operated face image is generated based on the operated waveform.

  In this way, the skin texture of the face image can be manipulated by manipulating at least one of the amplitude and frequency of the waveform representing the change in the gradation level of the face image. This makes it possible to change the age impression of the face image.

(2)
The operation means may change the texture of the skin of the face image by changing the amplitude of the waveform acquired by the acquisition means.

  In this case, the skin texture of the face image can be roughened by increasing the amplitude of the waveform representing the change in the gradation level of the face image. Thereby, the age impression of the face image can be increased. Further, the skin texture of the face image can be made finer by reducing the amplitude of the waveform representing the change in the gradation level of the face image. Thereby, the age impression of the face image can be lowered.

(3)
The operation means may change the amplitude of the waveform acquired by the acquisition means and change the gradation level of each point randomly within a predetermined range.

  In this case, the skin texture of the face image can be changed by changing the amplitude of the waveform representing the change in the gradation level of the face image. Due to the change in amplitude, a fine change in gradation level included in the original face image is lost. Therefore, by changing the gradation level of each point of the waveform randomly within a predetermined range, a minute change in gradation level included in the original face image is reproduced. Therefore, the age impression of the face image can be increased without degrading the image quality of the face image.

(4)
The operation means may change the texture of the skin of the face image by changing the frequency of the waveform acquired by the acquisition means.

  In this case, the skin texture of the face image can be roughened by lowering the frequency of the waveform representing the change in the gradation level of the face image. Thereby, the age impression of the face image can be increased. Further, the skin texture of the face image can be made finer by reducing the frequency of the waveform representing the change in the gradation level of the face image. Thereby, the age impression of the face image can be lowered.

(5)
The operation means substantially increases the frequency of the waveform by randomly changing the gradation level of each point of the waveform within a predetermined allowable range of the amplitude with reference to the change of the center level of the waveform acquired by the acquisition means. May be.

  In this case, the change in the center level of the waveform representing the change in the gradation level of the face image indicates a substantial change in the gradation level of the original face image. Therefore, by changing the tone level of each point of the waveform randomly within a predetermined allowable range of amplitude with reference to the center level of the waveform, the texture of the face image can be textured without greatly degrading the impression of the original face image. Can be made finer. Thereby, the age impression of the face image can be lowered without greatly impairing the impression of the face image.

(6)
The face image operation program according to the second invention is a face image operation program that can be executed by a computer. The face image operation program is stored in the storage means and processing for storing the human face image as data representing the gradation level. The computer is caused to execute a process for generating an age-operated face image by operating at least one of the amplitude and frequency of a waveform representing a change in the gradation level of the face image based on the obtained data.

  In the face image operation program according to the present invention, the human face image is stored as data representing the gradation level in the storage means. Based on the stored data, at least one of the amplitude and the frequency of the waveform representing the change in the gradation level of the face image is manipulated. Thereby, an age-operated face image is generated.

  In this way, the skin texture of the face image can be manipulated by manipulating at least one of the amplitude and frequency of the waveform representing the change in the gradation level of the face image. This makes it possible to change the age impression of the face image.

  According to the present invention, it is possible to manipulate the texture of the face image by manipulating at least one of the amplitude and frequency of the waveform representing the change in the gradation level of the face image. This makes it possible to change the age impression of the face image.

(1) Basic concept First, the basic concept of the face image operation method according to an embodiment of the present invention will be described.

  The skin is composed of fine irregularities, and the degree of the irregularities is perceived as fineness or roughness. From this, the unevenness of the skin is considered as a fine wave, and the skin texture is manipulated by manipulating the wave. Also, as a basic idea, the skin texture of adults is rougher than that of children.

  Consider frequency and amplitude as factors that define the degree of skin texture. FIG. 1 is a diagram showing the relationship between the frequency and amplitude of a waveform representing the skin texture of adults and children.

  Since a wave having a low frequency and a large amplitude is regarded as a rough wave, as shown by a solid line L in FIG. 1, the waveform representing the skin texture of an adult has a frequency that is higher than the waveform representing the skin texture of a child. Consider low and large amplitude.

  The texture of the skin of the face image can be represented by a change in gradation level in the skin area of the face image. Therefore, the gradation level change in the skin area of the face image is represented by a waveform, and the texture of the skin of the face image is manipulated by manipulating at least one of the frequency and amplitude of the waveform.

  The tone level of each pixel in the skin area of the face image is measured in advance for each age, and the waveform amplitude representing the change in tone level is obtained from the standard deviation of the tone level.

  For each age, the average value of the amplitude and frequency of the waveform representing the change in gradation level in the skin area of the face images of a plurality of persons is measured. In this way, the texture of the face image is controlled by manipulating at least one of the amplitude and frequency of the waveform based on the average value of the amplitude and frequency of the waveform representing the change in gradation level measured in advance for each age. It is possible to manipulate the age impression of the face image.

  Note that when the face image is displayed on the display device, the change in gradation level corresponds to the change in luminance. In this case, in order to manipulate the texture of the face image displayed on the display device, a waveform representing a change in the level of the luminance signal is manipulated.

  Since the face image is usually represented two-dimensionally, the gradation level in the skin area of the face image also changes two-dimensionally. In the following description, for easy understanding, a change in gradation level in the skin region of the face image is represented by a one-dimensional waveform. That is, the change in gradation level on an arbitrary straight line of the face image is represented by a waveform.

  In actual face image operations, the operations described below are performed in two orthogonal directions of the face image.

  In the following description, a waveform representing a change in gradation level in the skin area of the face image is referred to as a gradation waveform.

(2) Roughening of skin texture In order to roughen the skin texture of the face image, the amplitude of the gradation waveform is increased or the frequency is decreased.

  Actually, waveforms of different frequencies are mixed in a complicated manner for each position where the skin is focused. Therefore, when the frequency of the gradation waveform in a certain region is manipulated, it is necessary that the influence of the manipulation of the frequency does not affect other regions. Here, only the amplitude of the gradation waveform is operated in order to facilitate the operation of roughening the skin.

  FIG. 2 is a diagram for explaining the operation of the amplitude of the gradation waveform for roughening the skin texture. The vertical axis in FIG. 2 represents the gradation level of the face image, and the horizontal axis represents the position of the face image. In FIG. 2, the change in the gradation level of the face image is represented by a regular gradation waveform.

  In FIG. 2, the gradation waveform before the operation is represented by W0, and the gradation waveform after the operation is represented by W1. Further, the center level of the amplitude of the gradation waveform is represented by CL.

  As indicated by arrows in FIG. 2, the amplitude of the gradation waveform W0 is enlarged at a predetermined enlargement ratio to obtain a gradation waveform W1 having an enlarged amplitude. Thereby, the texture of the skin of the face image becomes rough.

  As described above, the average value of the amplitude of the gradation waveform is measured in advance for each age. In the aging operation of the face image, the amplitude enlargement ratio is obtained from the ratio of the amplitude of the gradation waveform corresponding to the age of the face image after the operation to the amplitude of the gradation waveform corresponding to the age of the face image before the operation.

  The change in the gradation level in the actual face image does not always change regularly as shown in FIG. FIG. 3 is a diagram illustrating an example of a change in the gradation level of an actual face image.

  As shown in FIG. 3, the center level CL of the gradation waveform W0 before the operation changes in a curved manner due to a change in skin color. Also in this case, the amplitude of the gradation waveform W0 is enlarged at a predetermined enlargement ratio to obtain a gradation waveform W1 with an enlarged amplitude. Thereby, the texture of the skin can be roughened while maintaining the tendency of the skin color change of the face image.

  When the amplitude of the gradation waveform is increased, an impression that the contrast of the skin has increased may be left, and unnaturalness may remain in the texture of the skin. One reason is that the waveform representing the change in gradation level in the skin of the actual facial image includes a plurality of waveforms having a small amplitude, but a waveform having a small amplitude can be obtained by enlarging the gradation waveform. Is thought to disappear. Therefore, a facial image that gives a natural impression is generated by adding a waveform with a small amplitude to the enlarged gradation waveform.

  FIG. 4 is a diagram showing a method of adding a small amplitude waveform to the gradation waveform.

  In FIG. 4, the dot of the gradation waveform W <b> 1 whose amplitude is enlarged indicates each pixel of the face image or each sampling point of the face image. The amplitude is changed vertically within a predetermined range based on random numbers for each dot of the gradation waveform W1. The change range of each dot is preferably a range from 0 to about the average expansion amplitude of the gradation waveform W1.

  Such a small amplitude waveform is a slight change in gradation level in the face image and cannot be perceived in detail as an impression of the face image. Therefore, by randomly adding a minute change to each dot of the gradation waveform W1, the skin texture can be roughened while maintaining the naturalness of the skin of the face image.

  In the present embodiment, the amplitude of the gradation waveform is enlarged in order to roughen the skin texture of the face image. However, the frequency of the gradation waveform may be lowered or the amplitude of the gradation waveform may be reduced. And the frequency of the gradation waveform may be lowered.

(3) Operation for making skin texture finer To make the skin texture of a face image finer, the amplitude of the gradation waveform is decreased or the frequency is increased.

(A) Operation for Reducing the Amplitude of the Grayscale Waveform FIG. 5 is a diagram for explaining the operation of the amplitude of the waveform for making the skin texture finer. The vertical axis in FIG. 5 represents the gradation level of the face image, and the horizontal axis represents the position of the face image. In FIG. 5, the change in the gradation level of the face image is represented by a regular gradation waveform.

  In FIG. 5, the gradation waveform before operation is represented by W0, and the gradation waveform after operation is represented by W1. Further, the center level of the amplitude of the gradation waveform is represented by CL.

  As indicated by an arrow in FIG. 5, the gradation waveform W0 is reduced in amplitude by a predetermined reduction ratio to obtain a gradation waveform W1 with reduced amplitude. Thereby, the skin texture of the face image becomes fine.

  As described above, the average value of the amplitude of the gradation waveform is measured for each age. In the face image rejuvenation operation, the amplitude reduction rate is obtained from the ratio of the amplitude of the gradation waveform corresponding to the age of the face image after the operation to the amplitude of the gradation waveform corresponding to the age of the face image before the operation.

  The change in the gradation level in the actual face image does not always change regularly as shown in FIG. FIG. 6 is a diagram illustrating an example of a change in the gradation level of an actual face image.

  As shown in FIG. 6, the center level CL of the gradation waveform W0 before the operation changes in a curved manner due to a change in skin color. In this case as well, the amplitude of the gradation waveform W0 is reduced at a predetermined reduction rate to obtain a gradation waveform W1 with reduced amplitude. Thereby, the texture of the skin can be made fine while maintaining the tendency of the skin color change of the face image.

(B) Operation for Increasing the Frequency of the Grayscale Waveform FIG. 7 is a diagram for explaining the operation of the frequency of the grayscale waveform for finer skin texture. The vertical axis in FIG. 7 represents the gradation level of the face image, and the horizontal axis represents the position of the face image. In FIG. 7, the change in the gradation level of the face image is represented by a regular gradation waveform.

  In FIG. 7, the dot of the gradation waveform W1 with reduced amplitude indicates each pixel of the face image or each sampling point of the face image. Here, for each dot of the gradation waveform W1 with reduced amplitude, the gradation level is randomly changed up and down within the range of the amplitude of the gradation waveform W1 based on a random number. Here, the range of the amplitude of the gradation waveform W1 whose amplitude is reduced is between the level TH1 and the level TH2.

  The frequency of the gradation waveform W1 can be substantially increased by randomly changing the gradation level vertically in the amplitude range for each dot of the gradation waveform W1.

  FIG. 8 is a diagram showing an example of the gradation waveform after the frequency operation. The vertical axis in FIG. 8 represents the gradation level of the face image, and the horizontal axis represents the position of the face image. In FIG. 8, the gradation waveform after the frequency operation is represented by W2.

  As shown in FIG. 8, the gradation waveform W2 after the frequency operation changes within the amplitude range (between level TH1 and level TH2) of the gradation waveform W1 of FIG. 7, and the gradation waveform of FIG. It has a substantially higher frequency than W1.

  Thus, since the amplitude of the gradation waveform W1 is sufficiently small, even if the gradation level of each dot of the gradation waveform W1 is changed within the amplitude range, it does not give a significantly different impression as a face image. The frequency is substantially increased.

  FIG. 9 is a diagram illustrating an example of a change in the gradation level of an actual face image. As shown in FIG. 9, the center level CL of the gradation waveform before the operation changes in a curved manner due to the change of the skin color. Also in this case, the frequency of the gradation waveform W1 can be substantially increased by randomly changing the gradation level up and down within the amplitude range for each dot of the gradation waveform W1 after the operation. Thereby, the texture of the skin can be made fine while maintaining the tendency of the skin color change of the face image.

  In the present embodiment, the amplitude of the gradation waveform is reduced and the frequency of the gradation waveform is increased in order to make the skin texture of the face image finer, but only one of them may be performed.

  In this embodiment, the gradation level is randomly changed up and down within the amplitude range for each dot of the gradation waveform after the amplitude reduction operation, but the gradation level is changed for each dot of the gradation waveform. May be changed randomly up and down exceeding the amplitude.

  Furthermore, the frequency of the gradation waveform may be increased by superimposing another waveform having a high frequency on the gradation waveform.

(4) Configuration of Face Image Manipulation Device FIG. 10 is a block diagram showing the configuration of the face image manipulation device according to the present embodiment.

  Next, the configuration of a face image manipulation device for executing the face image manipulation method of the present embodiment will be described with reference to FIG.

  The image processing apparatus 50 includes a CPU (Central Processing Unit) 501, a ROM (Read Only Memory) 502, a RAM (Random Access Memory) 503, an input device 504, a display device 505, an external storage device 506, a recording medium driving device 507, and the like. A printing device 508.

  The input device 504 includes a keyboard, a mouse, a scanner, a digital camera, and the like, and is used for inputting various commands, data, and images. The ROM 502 stores a system program. The recording medium driving device 507 includes a CD (compact disk) drive, a DVD (digital versatile disk) drive, an FD (flexible disk) drive, and the like, and reads / writes data from / to a recording medium 509 such as a CD, DVD, or FD. . In the recording medium 509, a face image operation program is recorded. The external storage device 506 includes a hard disk device or the like, and stores a face image operation program and various data read from the recording medium 509 via the recording medium driving device 507. The CPU 501 executes the face image operation program stored in the external storage device 506 on the RAM 503.

  The display device 505 includes a liquid crystal display panel, a CRT (cathode ray tube), and the like, and displays various images. The printing device 508 prints various images.

  As the recording medium 509 for recording the face image operation program, various recording media such as a semiconductor memory such as a ROM and a hard disk can be used. Further, the face image operation program may be downloaded to the external storage device 506 via a communication medium such as a communication line and executed on the RAM 503.

(5) Operation of Face Image Manipulating Device FIG. 11 is a flowchart showing the operation of the face image manipulating device of FIG. Hereinafter, the face image manipulation operation by the face image manipulation device of FIG. 10 will be described with reference to the flowchart of FIG.

  Here, a description will be given of a process of creating a face image having a different age of the subject by manipulating the face image of the subject. First, the operator inputs the face image data of the subject person and the age of the subject person through an input device 504 such as a digital camera or a keyboard.

  The CPU 501 stores face image data in the external storage device 506 (step S1). The face image data includes red image data, green image data, and blue image data. The following gradation waveform operations may be performed for each of the red image data, the green image data, and the blue image data, or may be performed on the red image data that affects the skin color.

  It is assumed that position information indicating the position of face parts (hair, eyes, mouth, etc.) in the face image is also acquired in advance and stored in the external storage device 506. The position information of the face part may be acquired by a human input operation or may be automatically acquired from the face image. Thereby, the CPU 501 extracts skin area data from the face image. Further, the external storage device 506 stores in advance an average value of amplitudes and an average value of frequencies of gradation waveforms of face images of each age.

  Next, the operator uses the input device 504 to designate whether to perform an aging operation or a aging operation, and inputs the age after the operation.

  The CPU 501 determines whether an aging operation is designated by the operator (step S2).

  When the aging operation is designated, the CPU 501 enlarges the amplitude of the gradation waveform of the face image stored in the external storage device 506 (step S3). Thereby, the texture of the skin of the face image becomes rough. In this case, the enlargement ratio is set based on the ratio of the average value of the amplitude of the gradation waveform corresponding to the age of the face image after the operation to the average value of the amplitude of the gradation waveform corresponding to the age of the face image before the operation. Is done.

  Next, the CPU 501 randomly adds a minute change for each dot of the gradation waveform by the method described with reference to FIG. 4 (step S4). Thereby, the texture of the skin can be roughened while maintaining the naturalness of the skin of the face image. As a result, the age impression of the face image can be aged.

  When the rejuvenation operation is designated in step S2, the CPU 501 reduces the amplitude of the gradation waveform of the face image stored in the external storage device 506 (step S5). Thereby, the skin texture of the face image becomes fine. In this case, the reduction ratio is set based on the ratio of the average value of the amplitude of the gradation waveform corresponding to the age of the face image after the operation to the average value of the amplitude of the gradation waveform corresponding to the age of the face image before the operation. Is done.

  Next, the CPU 501 randomly changes the gradation level for each dot of the gradation waveform within the amplitude range by the method described with reference to FIG. 7 (step S6). Thereby, the frequency of the gradation waveform is substantially increased.

  Thereby, the texture of the skin can be made fine while maintaining the naturalness of the skin of the face image. As a result, the age impression of the face image can be made younger.

  Finally, the CPU 501 displays the face image after the operation on the display device 505 and prints it with the printing device 508 (step S7).

(6) Example Actually, an operation for roughening the texture of the skin on the face image of the subject was performed. FIG. 12 is a diagram showing a face image of the subject person. FIG. 13 is a diagram showing the face image after the operation in the case where the operation of roughening the skin is performed on the face image of FIG.

  In this operation, the amplitude of the gradation waveform was enlarged at an enlargement ratio of 130%. Further, on the skin of the face image whose amplitude is enlarged, a minute change is randomly added to each dot of the gradation waveform by the method shown in FIG. As a result, the amplitude of the gradation waveform is enlarged, and minute changes are randomly added to each dot of the gradation waveform, so that the skin texture of the face image is roughened while maintaining the naturalness of the skin of the face image. I found out that I could do it.

  Next, the skin of the adult face image was compared with the skin of the child's face image, and an operation for making the skin texture of the adult face image finer was performed. In this operation, the amplitude of the gradation waveform was reduced at a reduction rate of 70%. Further, the tone level of each dot of the tone waveform was randomly changed up and down within the amplitude range by the method shown in FIG. 7 on the skin of the face image with reduced amplitude.

  FIG. 14 is an enlarged image of adult skin. FIG. 15 is a diagram showing an enlarged image after the above-described rejuvenation operation is performed on an image of an adult skin. FIG. 16 shows an enlarged image of the child's skin.

  From this result, the amplitude of the tone waveform is reduced, and the tone level of each dot of the tone waveform is randomly changed up and down within the amplitude range by the method shown in FIG. It was found that the skin texture of the face image can be made fine while maintaining the naturalness of the face.

(7) Correspondence between Each Component in Claim and Each Unit in Embodiment In this embodiment, the external storage device 506 corresponds to a storage unit, and the CPU 501 and the face image presentation program correspond to an operation unit.

  The present invention can be used for manipulating the age impression of a facial image, and can be used for various purposes such as criminal investigations.

It is a figure which shows the relationship between the frequency and amplitude of a waveform showing the skin texture of an adult and a child. It is a figure for demonstrating operation of the amplitude of the gradation waveform for roughening the skin texture. It is a figure which shows an example of the change of the gradation level of an actual face image. It is a figure which shows the method of adding the waveform of a small amplitude to a gradation waveform. It is a figure for demonstrating operation of the amplitude of the waveform for making skin texture fine. It is a figure which shows an example of the change of the gradation level of an actual face image. It is a figure for demonstrating the operation of the frequency of the gradation waveform for making skin texture fine. It is a figure which shows an example of the gradation waveform after the frequency operation. It is a figure which shows an example of the change of the gradation level of an actual face image. It is a block diagram which shows the structure of the face image operation apparatus which concerns on this Embodiment. It is a flowchart which shows operation | movement of the face image operation apparatus of FIG. It is a figure which shows a subject's face image. It is a figure which shows the face image after operation at the time of performing operation which roughens the skin texture to the face image of FIG. It is a figure which shows the enlarged image of adult skin. It is a figure which shows the enlarged image after operation at the time of performing said rejuvenation operation to the image of adult skin. It is a figure which shows the enlarged image of a child's skin.

Explanation of symbols

50 image processing apparatus 501 CPU
502 ROM
503 RAM
504 Input device 505 Display device 506 External storage device 507 Recording medium drive device 508 Printing device 509 Recording medium W0, W1, W2 Gradation waveform CL Center level

Claims (6)

A face image manipulation device for manipulating a face image,
Storage means for storing a human face image as data representing a gradation level;
Operating means for generating an age-operated face image by operating at least one of an amplitude and a frequency of a waveform representing a change in gradation level of the face image based on data stored in the storage means A face image operation device characterized by the above. 2. The face image operation device according to claim 1, wherein the operation means changes the texture of the skin of the face image by changing the amplitude of the waveform acquired by the acquisition means. 2. The face image operating device according to claim 1, wherein the operating means changes the amplitude of the waveform acquired by the acquiring means and randomly changes the gradation level of each point within a predetermined range. The face image manipulation device according to claim 1, wherein the operation unit changes a texture of the skin of the face image by changing a frequency of a waveform acquired by the acquisition unit. The operation means changes the frequency of the waveform by randomly changing the gray level of each point of the waveform within a predetermined allowable range of amplitude with reference to the change in the center level of the waveform acquired by the acquisition means. 5. The face image operating device according to claim 4, wherein the face image operating device is substantially high. A face image operation program executable by a computer,
A process of storing a human face image in the storage means as data representing a gradation level;
Processing to generate an age-operated face image by operating at least one of the amplitude and frequency of a waveform representing a change in the gradation level of the face image based on the data stored in the storage means. A face image operation program that is executed.

Images