Wrinkle detection device, wrinkle detection method, and wrinkle detection program

Abstract

To evaluate wrinkles easily.SOLUTION: A wrinkle detection device includes: an acquisition unit for acquiring a captured image of the face; a feature emphasis processing unit for generating a feature emphasis image by executing feature emphasis processing for a composite object image based on the captured image; an embossing processing unit for generating embossed images in a plurality of radiation directions by executing embossing processing when light is radiated to the feature emphasis image subjected to the feature emphasis processing in a pseudo manner from a plurality of irradiation directions; a compositing unit for generating image feature composite images in the plurality of irradiation directions by compositing the embossed images in the plurality of irradiation directions and a composite source image based on the captured image; a frequency component calculation unit for calculating a frequency component according to a change in a concentration value of each pixel by executing Fourier transformation for concentration value change information indicating the change in the concentration value of each pixel in the image feature composite images in the plurality of irradiation directions; and a control unit for executing control to output a wrinkle emphasis image in which wrinkles on the face are emphasized on the basis of the frequency component.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a wrinkle detection device, a wrinkle detection method, and a wrinkle detection program. [Background technology]

[0002] Conventionally, devices for detecting facial wrinkles have been proposed (see, for example, Patent Document 1). Methods for evaluating facial wrinkles include visual and photographic evaluation, replica methods, and in vivo methods.

[0003] The method of evaluation by visual inspection and photograph is a method in which wrinkles are scored by comparing wrinkles visually inspected by an expert or wrinkles in a photograph with standard photographs for each wrinkle grade.

[0004] The replica method involves using silicon to make a replica (mold) of the wrinkles, then shining light on the replica to generate a shadow to analyze the unevenness, or using a special device to irradiate the wrinkles with a laser and identify the unevenness from the distortion of the lines.The replica method makes it possible to calculate the depth, volume, and area ratio of wrinkles.

[0005] The in vivo method uses a specialized device to irradiate the wrinkles with a laser and identify the irregularities from the distortion of the lines. This method makes it possible to calculate the depth, volume, number, and area ratio of wrinkles. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-15118 Summary of the Invention [Problem to be solved by the invention]

[0007] However, methods for assessing wrinkles by visual inspection and photography require expert evaluation and are difficult to quantify. Furthermore, the replica method has problems such as requiring the subject to maintain the same facial expression, requiring skill to obtain replicas, making it difficult to analyze fine wrinkles, and being susceptible to the effects of air bubbles and sweat. Furthermore, the in vivo method has problems such as requiring the subject to maintain the same facial expression, requiring large and expensive equipment, and requiring skill to focus the laser.

[0008] All of these methods have problems such as placing a heavy burden on the subject and the evaluator, and basically only being able to evaluate wrinkles that are neutral.

[0009] An object of the present invention is to provide a wrinkle detection device, a wrinkle detection method, and a wrinkle detection program that enable easy wrinkle evaluation. [Means for solving the problem]

[0010] a feature enhancement processing unit that generates a feature-enhanced image by performing feature enhancement processing on a synthesis target image based on the captured image; an embossing processing unit that generates embossed images for the plurality of illumination directions by performing embossing processing on the feature-enhanced image when light is simulated from a plurality of illumination directions on the feature-enhanced image that has been subjected to the feature enhancement processing; a synthesis unit that generates an image feature synthesis image for the plurality of illumination directions by synthesizing the embossed images for the plurality of illumination directions with a source image based on the captured image; a frequency component calculation unit that calculates frequency components corresponding to changes in density value of each pixel by performing a Fourier transform on density value change information that represents changes in density value of each pixel in the image feature synthesis image for the plurality of illumination directions; and a control unit that controls to output a wrinkle-enhanced image in which wrinkles on the face are color-coded according to their depth based on the power of the frequency components.

[0011] In the wrinkle detection device according to the first aspect, the plurality of irradiation directions may be eight different directions, and the frequency component calculation unit may calculate frequency components at a spatial frequency of 2 / 8 (period / pixel) by performing a Fourier transform on density value change information of eight pixels, and the control unit may perform control based on the frequency components to output a wrinkle-enhanced image in which facial wrinkles are enhanced as a still image.

[0012] In the wrinkle detection device according to the first aspect, the feature enhancement processing unit may perform a smoothing process on the original synthesis image using a smoothing filter.

[0013] In the wrinkle detection device according to the first aspect, the plurality of irradiation directions may be eight different directions, and the frequency component calculation unit may calculate frequency components at a spatial frequency of 1 / 8 (period / pixel) by performing a Fourier transform on density value change information of eight pixels, and the control unit may perform control based on the frequency components to output a wrinkle-enhanced image in which the facial wrinkles have been binarized as a still image.

[0014] In the wrinkle detection device according to the first aspect, the control unit may calculate the wrinkle area ratio based on a binarized image in which the target region of the face is binarized.

[0015] In the wrinkle detection device according to the first aspect, the plurality of irradiation directions may be eight different directions, and the frequency component calculation unit may calculate frequency components at a spatial frequency of 1 / 8 (period / pixel) by performing a Fourier transform on density value change information of eight pixels, and the control unit may perform control based on the frequency components to output a wrinkle-enhanced image in which facial wrinkles are emphasized as a moving image.

[0016] A wrinkle detection method according to a second aspect includes a computer acquiring a photographed image of a face, performing feature enhancement processing on a synthesis target image based on the photographed image to generate a feature-enhanced image, performing embossing processing on the feature-enhanced image that has been subjected to feature enhancement processing to generate embossed images for the plurality of illumination directions by illuminating the feature-enhanced image with simulated light from a plurality of illumination directions, generating an image feature synthesized image for the plurality of illumination directions by synthesizing the embossed images for the plurality of illumination directions with a source image based on the photographed image, Fourier transforming density value change information representing changes in density value of each pixel in the image feature synthesized image for the plurality of illumination directions to calculate frequency components corresponding to the change in density value of each pixel, and performing control processing to output a wrinkle-enhanced image in which wrinkles on the face are emphasized based on the frequency components.

[0017] A wrinkle detection program according to a third aspect causes a computer to execute the following processes: acquire a photographed image of a face; generate a feature-enhanced image by performing feature enhancement processing on a synthesis target image based on the photographed image; generate embossed images for the plurality of illumination directions by performing embossing processing on the feature-enhanced image that has been subjected to feature enhancement processing in a manner similar to irradiating light from a plurality of illumination directions on the feature-enhanced image; generate an image feature composite image for the plurality of illumination directions by combining the embossed images for the plurality of illumination directions with a source image based on the photographed image; calculate frequency components corresponding to the change in density value of each pixel by Fourier transforming density value change information representing a change in density value of each pixel in the image feature composite image for the plurality of illumination directions; and output a wrinkle-enhanced image in which facial wrinkles are enhanced based on the frequency components. [Effects of the Invention]

[0018] According to the present invention, there is an effect that it is possible to easily evaluate wrinkles. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 2 is a configuration diagram showing a hardware configuration of a wrinkle detection device. [Figure 2] FIG. 2 is a block diagram showing the functional configuration of the wrinkle detection device. [Figure 3] FIG. 10 is a diagram for explaining the overall processing flow of wrinkle detection processing. [Figure 4] FIG. 10 is a diagram for explaining an irradiation direction. [Figure 5] FIG. 10 is a diagram for explaining density value change information. [Figure 6] FIG. 10 is a diagram illustrating a spatial frequency. [Figure 7] 10 is a flowchart showing an example of a processing flow of a wrinkle detection program. [Figure 8] FIG. 2 is a diagram illustrating an example of an original image. [Figure 9] FIG. 10 is a diagram showing an example of a wrinkle-enhancing image. [Figure 10] FIG. 10 is a diagram showing an example of a wrinkle-enhancing image. [Figure 11] FIG. 10 is a diagram showing an example of a wrinkle-enhancing image. [Figure 12] FIG. 10 is a diagram illustrating an example of a binarized image of a target region. [Figure 13] FIG. 1 shows replicas of the facial wrinkle area ratios of people in their 20s to 50s calculated using the replica method. [Figure 14] 1 is a graph showing the area ratio for each age group calculated using the replica method. [Figure 15] These are images of wrinkles of each age group when calculating the area ratio of facial wrinkles in people in their 20s to 50s using the in vivo method. [Figure 16] 1 is a graph showing the area ratio for each age group calculated using the in vivo method. [Figure 17] 10 shows binarized images of wrinkles of people in their 20s to 50s when the area ratio of facial wrinkles was calculated using the method according to this embodiment. [Figure 18] 10 is a graph showing the area ratio for each age group calculated using the method according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0020] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0021] Fig. 1 is a block diagram showing the hardware configuration of wrinkle detection device 10. As shown in Fig. 1, wrinkle detection device 10 includes controller 12. Controller 12 includes CPU (Central Processing Unit) 12A, ROM (Read Only Memory) 12B, RAM (Random Access Memory) 12C, and input / output interface (I / O) 12D. CPU 12A, ROM 12B, RAM 12C, and I / O 12D are connected to each other via system bus 12E. System bus 12E includes a control bus, an address bus, and a data bus.

[0022] Furthermore, a camera 14, an operation unit 16, a display unit 18, a communication unit 20, and a storage unit 22 are connected to the I / O 12D.

[0023] The camera 14 includes an imaging element such as a CCD (Charge Coupled Devices) image sensor or a CMOS (Complementary Metal Oxide Semiconductor) image sensor, and outputs a captured image of a subject to the controller 12.

[0024] The operation unit 16 includes, for example, a mouse and a keyboard.

[0025] The display unit 18 is configured with, for example, a liquid crystal display.

[0026] The communication unit 20 is an interface for performing data communication with an external device.

[0027] Storage unit 22 is configured with a nonvolatile external storage device such as a hard disk, and stores wrinkle detection program 22A, etc. CPU 12A loads wrinkle detection program 22A stored in storage unit 22 into RAM 12C and executes it.

[0028] Fig. 2 is a block diagram showing the functional configuration of CPU 12A of wrinkle detection device 10. As shown in Fig. 2, CPU 12A functionally includes acquisition unit 24, feature emphasis processing unit 26, embossing processing unit 28, synthesis unit 30, frequency component calculation unit 32, and control unit 34.

[0029] The acquisition unit 24 acquires a captured image of a face from the camera 14. In this embodiment, as shown in Fig. 3, a case will be described in which the captured image is an RGB color image 36 including an R image 36R, a G image 36G, and a B image 36B of the three primary colors of R (red), G (green), and B (blue).

[0030] 3, the feature emphasis processing unit 26 generates a feature-emphasized image 38 by performing feature emphasis processing on an R image 36R that is a synthesis target image based on the captured image acquired by the acquisition unit 24. In this embodiment, as shown in FIG. 3, a case will be described in which the synthesis target image is an R image 36R of an RGB color image, but the synthesis target image may also be a G image or a B image.

[0031] Specifically, the feature enhancement processing unit 26 performs known feature enhancement processing on the synthesis target image. Examples of known feature enhancement processing include the processing described in Reference 1 below, which performs texture enhancement processing, edge enhancement processing, etc.

[0032] (Reference document 1) Patent No. 6021053

[0033] In texture enhancement processing, for example, image processing is performed by applying a filter such as a correlation filter, a variance filter, a contrast filter, or an entropy filter to the synthesis target image. In edge enhancement processing, image processing is performed by applying a first-order differential Prewitt filter or a first-order differential template Robinson filter to the synthesis target image. Note that feature enhancement processing is not limited to texture enhancement processing and edge enhancement processing.

[0034] 3, the embossing processor 28 generates embossed images 40 with multiple illumination directions by embossing a feature-enhanced image 38, which is a composite target image that has been subjected to feature enhancement processing by the feature enhancement processor 26, in a manner similar to irradiating light from multiple illumination directions. Specifically, the embossing process is performed using a known method, such as the known method described in the above-mentioned Reference 1.

[0035] In this embodiment, the multiple irradiation directions are eight different directions as shown in FIG. 4, and the eight different irradiation directions are 0 degrees, 45 degrees, 90 degrees, 135 degrees, 180 degrees, 225 degrees, 270 degrees, and 315 degrees, which are shifted clockwise by 45 degrees from the top of the feature-emphasized image 38 as 0 degrees.

[0036] The embossing processor 28 uses filter coefficients expressed as a 3 × 3 matrix set for each irradiation direction to generate eight embossed images 40 embossed for when light is irradiated from each irradiation direction. Note that because the embossing process finds the difference between multiple pixels, when the embossing process is performed on the feature-emphasized image 38, the edges of the feature-emphasized image 38 are emphasized.

[0037] 3, the composition unit 30 generates an image feature composite image 42 by combining embossed images 40 for each of a plurality of irradiation directions, i.e., eight directions, generated by the embossing processing unit 28 with an original image based on a captured image. In this embodiment, a case will be described in which the original image is a captured image, i.e., an RGB color image 36. In this case, eight image feature composite images 42R are generated by combining eight embossed images 40 with an R image 36R, eight image feature composite images 42G are generated by combining eight embossed images 40 with a G image 36G, and eight image feature composite images 42B are generated by combining eight embossed images 40 with a B image 36B.

[0038] The frequency component calculation unit 32 performs a Fourier transform on the density value change information representing the change in density value of each pixel in the image feature composite image 42 of multiple illumination directions generated by the synthesis unit 30, and calculates frequency components corresponding to the change in density value of each pixel.

[0039] Specifically, as shown in Fig. 3, the frequency component calculation unit 32 generates density value change information 44R that represents changes in density value of each pixel in the image feature composite image 42R. Similarly, the frequency component calculation unit 32 generates density value change information 44G that represents changes in density value of each pixel in the image feature composite image 42G, and density value change information 44B that represents changes in density value of each pixel in the image feature composite image 42B.

[0040] An example of density value change information is shown in Fig. 5. For ease of explanation, Fig. 5 shows an example of the change in density value of the central pixel of 3 x 3 pixels that is part of the image feature composite image 42R. Note that the density value is expressed in 8 bits, for example, and can take a value from 0 to 255. In this embodiment, eight different directions are used as irradiation directions, so eight density values ​​are obtained for the same pixel.

[0041] As shown in Fig. 5, the density value of the central pixel changes from "100," "140," "160," "140," "100," "60," "40," and "60" for the eight irradiation directions: 0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°. Thus, the density value change information represents changes in density values, and is information representing the correspondence between multiple irradiation directions and density values. The frequency component calculation unit 32 calculates density value change information representing such changes in density values ​​for each pixel of each image feature composite image.

[0042] 3, the frequency component calculation unit 32 generates density value change information 44 by averaging the density value change information 44R, 44G, and 44B. That is, for each pixel of the density value change information 44R, 44G, and 44B, the density value change information 44 is generated by dividing the density value by 3.

[0043] Next, as shown in Fig. 3, frequency component calculation unit 32 performs a Fourier transform on density value change information 44 to calculate power spectrum information 46, which represents the power (amplitude) as a frequency component corresponding to the change in density value of each pixel. In this embodiment, the density change information for one pixel includes eight density values, so the power (amplitude) is calculated for each of four spatial frequencies (basis functions) f1 to f4 (f4>f1) shown in Fig. 6. Note that spatial frequencies f1 to f4 are 1 / 8, 2 / 8, 3 / 8, and 4 / 8 (period / pixel), respectively. The magnitude of this power corresponds to the depth of the wrinkles; the greater the power, the deeper the wrinkles, and the smaller the power, the shallower the wrinkles.

[0044] Control unit 34 controls output of a wrinkle-enhanced image in which facial wrinkles are enhanced based on the power at spatial frequency f2, i.e., the spatial frequency of 2 / 8 (cycle / pixel), among spatial frequencies f1 to f4 calculated by frequency component calculation unit 32. For example, control unit 34 generates a wrinkle-enhanced image by assigning a color to each pixel of RGB color image 36 acquired by acquisition unit 24 according to the magnitude of the power, and controls display unit 18 to display the image. That is, a wrinkle-enhanced image is displayed in a color-coded manner according to the depth of the wrinkles. This makes it possible to clearly grasp the wrinkles according to their depth. The generated wrinkle-enhanced image may be output to storage unit 22 for storage.

[0045] Next, the wrinkle detection process executed by the CPU 12A will be described with reference to the flowchart shown in FIG.

[0046] In step S100, the CPU 12A functions as the acquisition unit 24 and acquires from the camera 14 an RGB color image 36, which is an image of a human face captured by the camera 14.

[0047] In step S102, the CPU 12A functions as the feature enhancement processing unit 26 to generate a feature-enhanced image 38 by performing feature enhancement processing on the R image 36R, which is a synthesis target image based on the RGB color image 36 acquired in step S100.

[0048] In step S104, the CPU 12A, functioning as the embossing processor 28, performs embossing on the feature-emphasized image 38 generated in step S104, assuming that light is irradiated from eight different irradiation directions, thereby generating embossed images 40 corresponding to each of the eight irradiation directions. Note that in the embossing process, the degree of embossing, i.e., the degree of embossing, can be adjusted by adjusting an embossing coefficient, but in this embodiment, as an example, the embossing coefficient is set to the maximum value, which maximizes the degree of embossing.

[0049] In step S106, the CPU 12A, as the synthesis unit 30, generates image feature synthesis images 42R, 42G, and 42B by synthesizing the embossed images 40 for each of the eight irradiation directions generated in step S104 with the RGB color image 36, which is the source image.

[0050] In step S108, the CPU 12A, functioning as the frequency component calculation unit 32, generates density value change information 44R, 44G, and 44B representing changes in density value of each pixel in the image feature composite images 42R, 42G, and 42B generated in step S106, and averages these to generate density value change information 44. Then, a Fourier transform is performed on the density value change information 44. As a result, the power as a frequency component corresponding to the change in density value of each pixel is calculated for each of the spatial frequencies f1 to f4.

[0051] In step S110, CPU 12A, as control unit 34, controls display unit 18 to display, as a still image, a wrinkle-enhanced image in which facial wrinkles are emphasized based on the power of the frequency component of spatial frequency f2, among the frequency components corresponding to the change in density value of each pixel calculated for each of spatial frequencies f1 to f4 in step S108. That is, for example, for RGB color image 36, which is the original image as shown in FIG. 8, a wrinkle-enhanced image 48 as shown in FIG. 9 is generated by assigning a color to each pixel according to the magnitude of the power, and the generated image is displayed on display unit 18. This assigns different colors to the original image according to the depth of the wrinkles, making it possible to clearly grasp the wrinkles. Alternatively, upper and lower thresholds may be set within the range of possible powers, and wrinkle-enhanced image 48 may be generated by assigning colors only to powers in the range above the lower threshold and below the upper threshold. This allows only wrinkles of a desired depth to be displayed.

[0052] Incidentally, if wrinkle-enhanced image 48, in which facial wrinkles are emphasized based on the power of spatial frequency f1 instead of spatial frequency f2, is displayed as a still image on display unit 18, textures other than wrinkles will also be detected and displayed, making spatial frequency f1 unsuitable for detecting wrinkles. This is thought to be because skin has many irregularities such as pores and hairs, and using spatial frequency f1 would also extract these.

[0053] However, when wrinkle-enhanced image 48 is generated based on the power of spatial frequency f2, small details other than wrinkles may be displayed as noise, as shown in Fig. 9. Therefore, in step S102, CPU 12A, as feature enhancement processing unit 26, may perform smoothing processing using a smoothing filter on RGB color image 36, which is the source image, and then perform feature enhancement processing on smoothed R image 36R, thereby generating feature-enhanced image 38. This makes it possible to display wrinkle-enhanced image 48, in which almost only wrinkles are emphasized, as shown in Fig. 10. It is also possible to allow the user to set whether or not to perform smoothing processing.

[0054] 9, different colors are displayed for each wrinkle depth. However, if you want to display all wrinkles regardless of their depth, it may be better to generate wrinkle-enhanced image 48 based on the power of spatial frequency f1. In this case, both the upper and lower thresholds are set to "0." This makes the wrinkles less noticeable, and wrinkle-enhanced image 48 is obtained, as shown in FIG. 11, in which the wrinkles are emphasized in a color representing less noticeable wrinkles (e.g., red).

[0055] Note that, when it is desired to calculate the wrinkle area ratio, i.e., the wrinkle area per unit area, it is possible to generate binarized image 50 in which regions of a color other than the color of interest (e.g., red) in the target region for which the area ratio is to be calculated are colored white and regions of colors other than the color of interest are colored black, and then calculate the wrinkle area ratio based on the generated binarized image, as shown in Fig. 12. Specifically, the area of ​​the white region in binarized image 50 is calculated, and the wrinkle area ratio can be calculated by dividing the calculated area of ​​the white region by the area of ​​the entire binarized image 50.

[0056] Here, Figure 13 shows replicas of the facial wrinkle area ratios for people in their 20s, 30s, 40s, and 50s calculated using the replica method, and Figure 14 shows a graph showing the area ratios for each age group calculated using the replica method based on the replicas of Figure 13.

[0057] FIG. 15 shows images of wrinkles taken at each age when the area ratio of facial wrinkles at each age was calculated using the in vivo method, and FIG. 16 shows a graph showing the area ratio at each age calculated using the in vivo method based on the images in FIG. 15.

[0058] FIG. 17 shows a binarized image obtained by binarizing the target area of ​​an image of wrinkles of each age group when the area ratio of facial wrinkles of each age group was calculated using the method according to the present embodiment described above, and FIG. 18 shows a graph showing the area ratio of each age group calculated using the method according to the present embodiment based on the binarized image of FIG. 17.

[0059] As shown in FIG. 18, it was found that the method according to this embodiment can detect shallow wrinkles in people in their 30s with higher accuracy than the replica method and the in vivo method.

[0060] Furthermore, although wrinkle-enhanced image 48 is displayed on display unit 18 as a still image in step S110, it may also be displayed as a moving image. That is, the captured images acquired in step S100 may be converted into a moving image, and the processing of steps S102 to S110 may be performed on each frame image constituting the moving image. In this case, if wrinkle-enhanced image 48 that has been smoothed in step S102 is generated and displayed as a moving image, the processing load may be heavy and the image may not be displayed smoothly. In this case, as described above, wrinkle-enhanced image 48 may be generated based on the power of spatial frequency f1 by setting both the upper and lower thresholds to "0" and displayed as a moving image. This allows wrinkle-enhanced image 48 to be displayed as a moving image smoothly.

[0061] As described above, in this embodiment, density value change information representing changes in density value of each pixel in an image with image features synthesized from a plurality of irradiation directions is Fourier transformed to calculate frequency components corresponding to the changes in density value of each pixel, and a wrinkle-enhanced image in which facial wrinkles are emphasized is displayed based on the calculated frequency components, thereby making it possible to easily evaluate wrinkles.

[0062] Furthermore, it is now possible to evaluate "expression wrinkles," which were previously impossible to evaluate. "Expression wrinkles" refer to facial wrinkles whose depth and shape change with facial or facial expression movements. Conventional evaluation methods target wrinkles when the subject is expressionless, and require the subject to maintain the same facial expression, making evaluation of expression wrinkles extremely difficult. In contrast, the present invention makes it possible to evaluate dynamic wrinkles using moving images or in real time, thereby enabling evaluation of expression wrinkles.

[0063] Although the embodiments have been described above, the technical scope of the present invention is not limited to the scope described in the above embodiments. Various modifications or improvements can be made to the above embodiments without departing from the gist of the invention, and such modifications or improvements are also included in the technical scope of the present invention. For example, the present invention can be applied to evaluating wrinkles on the neck and back of the hand in addition to facial wrinkles.

[0064] In the present embodiment, wrinkle detection program 22A is installed in storage unit 22, but the present invention is not limited to this. Wrinkle detection program 22A according to the present embodiment may be provided in a form recorded on a computer-readable storage medium. For example, wrinkle detection program 22A according to the present embodiment may be provided in a form recorded on an optical disc such as a CD (Compact Disc)-ROM or a DVD (Digital Versatile Disc)-ROM, or in a semiconductor memory such as a USB (Universal Serial Bus) memory or a memory card. Furthermore, wrinkle detection program 22A according to the present embodiment may be acquired from an external device via a communication line connected to communication unit 20. [Explanation of symbols]

[0065] 10 Wrinkle detection device 12 Controllers 14 Camera 16 Control section 18 Display 20 Communications Department 22 Memory section 22A Wrinkle Detection Program 24 Acquisition Department 26 Feature emphasis processing section 28 Embossing processing section 30 Synthesis section 32 Frequency component calculation section 34 Control Unit 36 RGB color images 38 Feature-enhanced images 40 Embossed Images 42 Image feature synthesis image 44 Density value change information 46 Power Spectrum Information 48 Wrinkle-enhancing images 50 Binarized Images

Claims

1. an acquisition unit that acquires a captured image of a face; a feature emphasis processing unit that generates a feature emphasized image by performing feature emphasis processing on a synthesis target image based on the captured image; an embossing processing unit that performs embossing processing on the feature-enhanced image by irradiating the image with light from eight different irradiation directions, each of which is shifted by 45 degrees, to generate embossed images of the eight different irradiation directions; a synthesis unit that generates a synthesis image of the eight different illumination directions by synthesizing the embossed images of the eight different illumination directions with a synthesis source image based on the photographed image; a frequency component calculation unit that calculates frequency components corresponding to the change in density value of each pixel by Fourier transforming the density value change information representing a change in density value of each pixel in the image feature composite image of the eight different irradiation directions, the density value change information being the density values ​​of the eight different irradiation directions of each pixel; a control unit that controls the output of a wrinkle-enhanced image in which the facial wrinkles are enhanced based on the power of the frequency components; Equipped with the frequency component calculation unit calculates frequency components at a spatial frequency of 2 / 8 (period / pixel) by performing a Fourier transform on the density value change information, and the control unit controls to output a wrinkle-enhanced image in which the facial wrinkles are enhanced as a still image based on the frequency components, thereby outputting a wrinkle-enhanced image in which the facial wrinkles are color-coded according to their depth, and The frequency component calculation unit calculates frequency components at a spatial frequency of 1 / 8 (cycle / pixel) by performing a Fourier transform on the density value change information, and the control unit controls to output a wrinkle-enhanced image in which the facial wrinkles have been binarized as a still image based on the frequency components, thereby controlling to output a wrinkle-enhanced image in which all wrinkles are enhanced regardless of the depth of the wrinkles. Wrinkle detection device.

2. The feature emphasis processing unit performs a smoothing process on the original image using a smoothing filter. The wrinkle detection device according to claim 1.

3. The control unit calculates the wrinkle area ratio based on a binarized image in which the target region of the face is binarized. The wrinkle detection device according to claim 1.

4. The computer Acquire a photographed image of the face, generating a feature-enhanced image by performing feature enhancement processing on a synthesis target image based on the photographed image; generating embossed images of eight different irradiation directions by irradiating the feature-emphasized image with light from eight different irradiation directions, each of which is shifted by 45 degrees; generating an image feature composite image of the eight different illumination directions by synthesizing the embossed images of the eight different illumination directions with a synthesis source image based on the photographed image, density value change information representing a change in density value of each pixel in the image feature composite image of the eight different illumination directions, wherein the density change information, which is the density value of each pixel in the eight different illumination directions, is Fourier transformed to calculate a frequency component corresponding to the change in density value of each pixel; controlling the output of a wrinkle-enhanced image in which the facial wrinkles are enhanced based on the power of the frequency components; a Fourier transform is performed on the density value change information to calculate frequency components at a spatial frequency of 2 / 8 (cycle / pixel), and a wrinkle-enhanced image in which the facial wrinkles are enhanced is output as a still image based on the frequency components, thereby outputting a wrinkle-enhanced image in which the facial wrinkles are color-coded according to their depth; and The density value change information is subjected to a Fourier transform to calculate frequency components at a spatial frequency of 1 / 8 (cycle / pixel), and a wrinkle-enhanced image in which the facial wrinkles are binarized based on the frequency components is controlled to be output as a still image, thereby controlling to output a wrinkle-enhanced image in which all wrinkles are enhanced regardless of the depth of the wrinkles. A wrinkle detection method for performing the process.

5. On the computer, Acquire a photographed image of the face, generating a feature-enhanced image by performing feature enhancement processing on a synthesis target image based on the photographed image; generating embossed images of eight different irradiation directions by irradiating the feature-emphasized image with light from eight different irradiation directions, each of which is shifted by 45 degrees; generating an image feature composite image of the eight different illumination directions by synthesizing the embossed images of the eight different illumination directions with a synthesis source image based on the photographed image, density value change information representing a change in density value of each pixel in the image feature composite image of the eight different illumination directions, wherein the density change information, which is the density value of each pixel in the eight different illumination directions, is Fourier transformed to calculate a frequency component corresponding to the change in density value of each pixel; controlling the output of a wrinkle-enhanced image in which the facial wrinkles are enhanced based on the power of the frequency components; a Fourier transform is performed on the density value change information to calculate frequency components at a spatial frequency of 2 / 8 (cycle / pixel), and a wrinkle-enhanced image in which the facial wrinkles are enhanced is output as a still image based on the frequency components, thereby outputting a wrinkle-enhanced image in which the facial wrinkles are color-coded according to their depth; and The density value change information is subjected to a Fourier transform to calculate frequency components at a spatial frequency of 1 / 8 (cycle / pixel), and a wrinkle-enhanced image in which the facial wrinkles are binarized based on the frequency components is controlled to be output as a still image, thereby controlling to output a wrinkle-enhanced image in which all wrinkles are enhanced regardless of the depth of the wrinkles. A wrinkle detection program that performs the process.

Images