KR101908785B1 - Tongue region extraction method and image processing apparatus for performing the method - Google Patents

Abstract

The present invention relates to a tongue region extraction method and an image processing apparatus that performs the method.
The tongue region extracting method includes acquiring an image including a user's tongue using a camera mounted on the image processing apparatus, deleting a skin region excluding a tongue region included in the image, extracting an outline of the tongue region, Based on the extracted outline, we set the center line of the tongue area, divide the tongue area at regular intervals, and apply the curvature calculation to the tongue boundary of the divided tongue area to correct the tongue area, extracting the tongue area from the corrected image can do.

Description

TECHNICAL FIELD [0001] The present invention relates to a tongue region extracting method and an image processing apparatus for performing the method. [0002]

The following description relates to an image processing technique for detecting a tongue area from an image, and more particularly to a method for extracting a tongue area more effectively in consideration of a flash effect applied to an image acquired through a camera mounted on a mobile phone will be.

In the conventional case, since the tongue image has characteristics similar to those of surrounding colors over the lip-tongue-gum, it is difficult to extract only the tongue area in the image by using the existing image extraction method alone.

Especially, when the tongue area is acquired through the camera mounted on the mobile, the surrounding environment such as brightness and color is not controlled in the tongue area. In addition, since the flash that is momentarily tapped to offset the camera's ambient brightness value is positioned next to the lens, a shadow is generated around the tongue in a specific direction, and at the same time, the boundary of the tongue portion directly becomes blurred.

Therefore, the conventional tongue region extraction algorithm has a problem that it is difficult to extract an accurate tongue region from a captured image in a mobile environment.

The present invention can provide a tongue area extraction method and an image processing apparatus that can more effectively extract a tongue area included in an image acquired in a mobile environment where the shooting environment is not constant.

The present invention can provide a tongue area extraction method and an image processing apparatus that can increase the accuracy of extracting a tongue area in consideration of a flash effect applied to an image in the process of shooting an image including a tongue area.

The tongue image extracting method performed by the image processing apparatus according to an exemplary embodiment includes the steps of acquiring an image including a user's tongue using a camera mounted on the image processing apparatus, box-; Deleting a skin area excluding a tongue area included in the image; Extracting an active contour for the tongue region from the images of the skin region deleted; Setting a center line of the tongue area based on the extracted outline, and dividing the tongue area at regular intervals; Correcting the tongue area by applying curvature calculation to the tongue interface for each of the divided tongue areas; And extracting a tongue region from the corrected image.

According to an embodiment, the deleting step may delete the skin area included in the image using colorimetric information (CIE a Thresholding) and color information (Hue Thresholding).

According to an exemplary embodiment of the present invention, the step of deleting the skin region may include applying a median filtering process to remove noise in the deleted skin region based on the threshold value of the color information, Can be performed.

According to an exemplary embodiment, the threshold value of the color information may be used to further remove the skin region that has not been deleted in the tongue region from which noise has been removed.

According to an exemplary embodiment, the dividing may include: calculating a length in the top direction of the tongue region according to the outline; dividing the calculated length in the upward direction into uniformly spaced intervals; can do.

According to an embodiment, the step of correcting may include dividing a loupe area according to the length of the loupe area in the up and down direction, and performing a linear correction with the center points of the divided lines to determine a center line of the tongue; Setting a final centerline for the entire tongue area based on the determined centerline of the tongue; And comparing the length between the left eye and the right eye of each region of the entire tongue region with respect to the determined final center line, and applying an image opening according to the error rate between the lengths.

According to an exemplary embodiment of the present invention, the correction of the skin region and the tongue region in the image may be performed through the ROI flattening (image opening) due to the flash effect generated in the process of acquiring the image based on the outline have.

The step of correcting according to an exemplary embodiment may be performed in consideration of a flash effect generated in the process of acquiring the image, i) a tongue area which is mistakenly recognized as a skin area in an image, or ii) Correction can be performed on the non-skin area.

The image processing apparatus according to an exemplary embodiment of the present invention predicts the position of a flash through header information of an image acquiring unit-image acquiring an image including a user's tongue using a camera mounted on the image processing apparatus; A skin region deletion unit for deleting a skin region excluding a tongue region included in the image; An outline extracting unit for extracting an outline of the tongue area from the images of the skin area deleted; A tongue region dividing unit that sets a center line of the tongue region based on the extracted outline and divides the tongue region at regular intervals; A tongue area correcting unit for correcting a tongue area by applying a curvature calculation to the tongue interface for each divided tongue area; And a tongue region extracting unit for extracting a tongue region from the corrected image.

The skin region deletion unit may delete the skin region included in the image using colorimetric information (CIE a Thresholding) and color information (Hue Thresholding).

The skin region deletion unit may perform a preprocessing operation on the skin region deleted from the image by applying median filtering to remove noise in the deleted image of the skin region based on the threshold value of the colorimetric information have.

The skin region deletion unit according to an exemplary embodiment may further remove the skin region that has not been deleted in the tongue region from which noise has been removed by using the threshold value of the color information.

The tongue region dividing unit according to an embodiment calculates the length in the top direction of the tongue region according to the outline, divides the calculated length in the upward direction at a uniform interval, and the divided image includes the tongue region in the tongue region can do.

The tongue area correcting unit may include a center line determining unit for dividing a lounging area according to the length of the louver area in the vertical direction and performing linear correction with the center points of the divided lines to determine an initial center line of the tongue; A final centerline setting unit for setting a final centerline for the entire tongue area based on the determined centerline of the tongue; And a flattening application unit for applying an image opening according to the error rate between the lengths of the left and right eyes of the respective regions of the entire tongue region based on the determined final center line.

The tongue area correcting unit according to an embodiment may be configured to perform a tongue area correcting operation based on i) a tongue area erroneously recognized as a skin area in an image, or ii) a tongue area erroneously recognized as a tongue area in an image, Correction can be performed on the non-skin area.

The tongue region extraction method and the image processing apparatus according to an embodiment of the present invention use a CIE a and a Hue value from an image to exclude a skin region (lip, gums, etc.) other than a tongue region, By extracting the tongue area, a more elaborate tongue area can be extracted.

The tongue region extraction method and the image processing apparatus according to an embodiment of the present invention divide the boundary between the tongue region and the skin region and the portion where the shadow is formed in consideration of the flash effect applied in the process of acquiring the image , And extraction accuracy of the tongue area can be increased by extracting the tongue area.

1 is a conceptual diagram for acquiring an image including a tongue area through a camera mounted on a mobile phone according to an embodiment.
2 is a view for explaining an image to which a flash effect is applied according to an embodiment.
3 is a diagram for explaining a detailed configuration of an image processing apparatus for extracting a tongue area according to an embodiment.
FIG. 4 is a diagram for explaining a process of deriving a final centerline of a tongue region with reference to a lounging region according to an embodiment and detecting a tongue region according to the final centerline.
FIG. 5 is a diagram illustrating a result of acquiring a tongue area from an image according to an embodiment.

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

1 is a conceptual diagram for acquiring an image including a tongue area through a camera mounted on a mobile phone according to an embodiment.

Referring to FIG. 1, the present invention can acquire an image 102 through a camera 104 mounted on a smartphone in a mobile environment. Here, the image 102 is an image including the tongue area of the user. For this purpose, the user can pull out the tongue to the maximum extent and take a stopped pose, and the image processing apparatus 101 can recognize the user's tongue An image including a tongue area can be obtained. Here, the image processing apparatus 101 may be an apparatus capable of acquiring an image including a tongue area of a user.

Specifically, the image processing apparatus 101 can acquire an image including a tongue area of a user by using a camera mounted on the image processing apparatus. At this time, the image processing apparatus is a smart phone which is easy to carry, and the camera may be mounted on a smart phone. Then, the image processing apparatus can acquire an image including not only the tongue area but also the skin area around the tongue.

The image processing apparatus 101 can acquire an image including a tongue area in a state where the distance between the camera and the tongue is arranged close to each other due to the characteristics of the mobile environment. Since the image processing apparatus can acquire an image without regard to time and place, it can be influenced by the external environment at the time of acquiring the image. In addition, when the effect of the flash 105 of the smartphone is utilized, a shadow due to light is generated in the tongue portion of the user, thereby acquiring an image having an ambiguous boundary between the tongue and another region.

In other words, when acquiring an image including a tongue area of a user through a smart phone, as the flash 105 of a general smart phone is positioned in the right direction from the front of the user, when the flash is operated near the tongue, The brightness of the surroundings is canceled, and the boundary between the tongue area and the skin area (lip / gums, etc.) is blurred on the left side of the image, and the shadow on the right side of the image protrudes to the outside.

In this case, a part of the left lips, whose boundary is ambiguous around the tongue area in the image, may be mistaken as a part of the tongue area. In order to correct this, an image opening process is performed.

Accordingly, in extracting a tongue area from an image including a tongue area to which a flash effect is applied, it is possible to suggest a method of extracting a more accurate tongue area in consideration of various external environments at the time of acquiring the image have.

2 is a view for explaining an image to which a flash effect is applied according to an embodiment.

Referring to FIG. 2, the image includes a tongue region and a skin region to which a flash effect is applied. At this time, the image may be divided into an area where the boundary is blurred and a region where the shadow is generated, depending on the position of the flash attached to the smartphone for photographing the image. For example, if a flash is positioned on the right side of a camera based on a camera mounted on a smart phone, a boundary region between the tongue region and the skin region occurs in the left region of the image and a shadow region occurs in the right region of the image . As another example, when a flash is located on the left side of the camera, a boundary obscure area and a shadow area occur in a direction opposite to the above-mentioned condition.

That is, since the position of the flash is biased to one side to the left or the right with respect to the camera, the boundary between the tongue area and the skin area is blurred in the portion directly irradiated with the flash, And a shadow is generated between the skin area, and the boundary between the two is clearly distinguished. In other words, when photographing the tongue area through a smart phone, the illumination due to the flash effect is not uniformly illuminated, resulting in: I) the boundary between the tongue area and the lips becomes blurred, and the lips are mistaken for the tongue area Occurs.

Therefore, in consideration of the flash effect applied to the image, the present invention can be applied to a region where the boundary between the tongue region and the skin region is ambiguous as in the region (a) of FIG. 2, Similarly, after setting the centerline of the tongue area based on the area where the shadows are generated, the tongue area is corrected by applying the curvature calculation to the tongue area of the tongue area divided at regular intervals, Tongue area can be extracted.

Here, the region (c) of FIG. 2 is a lingual region of the tongue, and as shadows occur in the lower portion of the lingual portion as shown in the drawing, it may be easy to set a more standard to measure the total length of the tongue.

3 is a diagram for explaining a detailed configuration of an image processing apparatus for extracting a tongue area according to an embodiment.

Referring to FIG. 3, the image processing apparatus 301 can perform an operation for more accurately extracting a tongue area in an image to which a flash effect is applied. To this end, the image processing apparatus includes an image acquiring unit 302, a skin region deleting unit 303, an outline extracting unit 304, a tongue region dividing unit 305, a tongue region correcting unit 306, 307).

The image acquisition unit 302 may acquire an image including a tongue area of the user through a camera mounted on the image processing apparatus or interlocked with the image processing apparatus. At this time, the image processing apparatus can acquire an image including not only the tongue area of the user but also the skin area of the user. Here, the skin area of the user may include lips, gums, etc. located adjacent to the tongue.

The skin region deleting unit 303 may delete the skin region included in the image using colorimetric information (CIE a Thresholding) and color information (Hue Thresholding). In other words, the skin area deleting unit 303 can primarily delete the skin area included in the image using the threshold value of the colorimetric information. Then, the skin area deleting unit 303 can delete the skin area primarily for the image in which the skin area has been deleted primarily using the threshold value of the color information. The present invention can delete the area of the lips, gums, and the like from a variety of information by deleting the skin area in the image through each reference (threshold value) using the colorimetric information and the color information.

The outline extracting unit 304 can extract an outline (active contour) for the tongue area from the deleted images of the skin area. Thus, the extracted outline can be utilized as information for correcting the skin region and the tongue region in the image.

The tongue area dividing unit 305 can divide the tongue area in the image so as to have a predetermined size based on the outline extracted through the outline extracting unit 304. [ Specifically, when acquiring an image, when the tongue is photographed at the center of the image, the position of the lens (camera) and the flash is naturally located above the glare (term of the tongue) A shadow in the downward direction can be formed at the interface of the losing region.

In addition, since the image is taken by the user as long as the tongue is extended, the lips having the most similar color to the tongue, especially the lower lip, are obscured by the tongue. Due to this, due to the lower lips obscured by the tongue and the shadow formed below the lacrimal region, the laryngeal region becomes clear of the boundary between the surrounding skin regions, so that the region can be more reliably extracted.

Based on the interest in the image, the tongue area dividing unit 305 can calculate the length of the tongue area in the vertical direction after applying the active contour algorithm, i.e., extracting the outline of the tongue area. Then, the tongue area dividing unit 305 can divide the tongue area based on the length of the calculated tongue area to obtain the lingual area corresponding to the end of the tongue.

Then, the tongue area dividing unit 305 can divide the gum area obtained by dividing the tongue area into three equal parts in the up and down directions. The tongue area dividing unit 305 can calculate the initial centerline of the tongue by applying linear interpolation to the center points of the trisected lines.

Thereafter, the tongue region dividing section 305 determines a final centerline for the tongue region by drawing a new line in a direction perpendicular to the calculated initial centerline, and divides the tongue region into five vertically, and the tongue region is based on the determined final centerline It is possible to divide the tongue region into 10 regions in total.

The tongue area correcting unit 306 can measure the distance from the final center line to the left end and the length to the right end for each region and compares the measured distances with respect to the tongue region divided into a total of 10 regions. The tongue region correcting unit 306 can correct a region that is misjudged to the lips beyond the tongue boundary by setting the size of the mask differently in the image open process according to the error rates of the lengths of the left and right limits. A more detailed operation will be described with reference to FIG.

At this time, according to the present invention, by performing the above-described method, since the strong flattening is performed in the tongue-depressed portion and the weak flattening is performed in the remaining portion, the loss of the existing tongue region due to the opening is reduced, Extraction may be possible.

The tongue area extracting unit 307 can extract an image including the finally corrected tongue area.

FIG. 4 is a diagram for explaining a process of deriving a final centerline of a tongue region with reference to a lounging region according to an embodiment and detecting a tongue region according to the final centerline.

Referring to FIG. 4 (a), the image processing apparatus can acquire an image including a tongue area to which a flash effect is applied.

Referring to FIG. 4B, the image processing apparatus can calculate the length in the up and down direction of the tongue area obtained roughly by the active contour algorithm (algorithm for extracting the outline). Then, the image processing apparatus can divide the tongue area into five equal parts based on the calculated length in the vertical direction, and set the loser area (the end area of the tongue). At this time, due to the flash effect applied to the image, the present invention makes it possible to extract the lure portion more accurately since the boundary between the larynx and the surrounding skin region becomes clear without the region misunderstood as the lips. In other words, according to the present invention, since the user lids the tongue in the process of acquiring the image, the lower lip of the user is covered, so that the probability of the lower lip occluded by the tongue to be mistaken as the tongue is rare in the lounging area, .

 Referring to FIG. 4 (c), the image processing apparatus further divides the total length of the laryngeal region into three equal parts to connect the center points at 0, 1/3, 2/3, and 3/3 points from the tongue end, The cubic equation can be calculated by linear interpolation.

Then, the image processing apparatus corrects the shadowed area formed by the flash in the image and corrects the angle of the final center line based on the calculated initial center line in the linear interpolation method.

Referring to FIG. 4 (d), the image processing apparatus sets a boundary line in the horizontal direction of the final centerline, performs re-segmentation of the divided tongue region after performing the active contour algorithm, and calculates curvature along the tongue boundary in each region . In other words, according to the present invention, the length of the width of each re-divided area can be compared to determine the symmetry. In addition, the present invention corrects an area for a boundary and an area where a shadow is generated. By adjusting the mask size and performing imgee opening for an area in which a difference in curvature is large compared to a neighboring boundary, Obtain the tongue area.

That is, the image processing apparatus can exclude areas that are mistaken for tongue due to the flash, and ROI can be flattened (Image opening). Thereafter, the image processing apparatus can extract the tongue region from the corrected image.

 In the present invention, by adjusting the size of the msak, the tongue area and the skin area are adjusted so that strong planarization is performed in the part that is mistaken as a tongue, and weak planarization is performed in the remaining part. That is, instead of applying a flattening technique of the same size throughout the tongue region, the flattening technique of different sizes is applied according to the characteristics of the respective regions, so that the loss of the tongue region is reduced by the image opening conventionally, It becomes possible to extract the complete tongue area as shown in Fig. 4 (e).

5 is an overall flow diagram for extracting tongue regions according to one embodiment.

In step 501, the image processing apparatus can acquire an image including a tongue through a camera. In detail, the image processing apparatus can acquire a user's image by using a camera mounted on the image processing apparatus. In order to more accurately determine the flash effect applied to the image, the present invention searches for a header file of the image to determine the type of the smartphone (camera) that captured the image and the position where the flash effect occurred, have.

The image processing apparatus can perform thresholding & morphological processing on the image. More specifically, the image includes a skin area excluding the tongue area and the tongue area, and the skin area may be a body part adjacent to the user's tongue, and may include lips, gums, and the like.

In step 502, the image processing apparatus can delete the skin region included in the image using the colorimetric information (CIE a Thresholding). The image processing apparatus can delete the skin region included in the image using the threshold value of the colorimetric information.

More specifically, the image processing apparatus can perform a preprocessing operation to fill a snowy region removed by the thresholding based on the condition according to Equation 1 as follows.

[Equation 1]

As a result, the image processing apparatus can perform as much as extracting the tongue images captured in the controlled environment by canceling the effects of the shadow and light reflection through the preprocessing process and extracting from the improved image.

In step 503, the image processing apparatus can remove the noise included in the image by applying median filtering to the image in which the skin region is deleted. In other words, the image processing apparatus can perform an operation to fill a vacant space generated in a tongue area, i.e., a snowy area, excluding a deleted skin area in an image based on a threshold value of colorimetric information.

In step 504, the image processing apparatus can perform a morphological process for crisis of the empty space for the image.

In step 505, the image processing apparatus may perform an additional deletion operation on the deleted skin region using the threshold value of hue thresholding. In other words, the image processing device can perform an additional deletion of the skin area related to the lips and gums in the image where the skin area is deleted. Here, the image processing apparatus can perform an addition deletion operation based on the condition according to Equation (2) below.

[Equation 2]

In step 506, the image processing apparatus may perform a morphological process for crisis of a blank space for an image.

In step 507, the image processing apparatus can extract an active contour for the tongue area from among the deleted images of the skin area.

Thereafter, the image processing apparatus can perform post-processing for correcting the skin region and the tongue region in the image based on the extracted outline.

In step 508, the image processing apparatus may divide an area including the losing area in the image. In other words, the image processing apparatus can divide the entire region of the tongue into five equal parts based on the vertical length of the tongue area. Then, the image processing apparatus can select the laryngeal region among the five divided regions, divide the laryngeal region into three, and set the initial centerline for the laryngeal region.

In step 509, the image processing apparatus can set the final center line of the entire tongue area along the initial center line based on the width of the loupe area.

In step 510, the image processing apparatus may re-divide the tongue area based on the final centerline of the set tongue area. That is, the image processing apparatus can divide the tongue region into 10 regions by dividing the entire tongue region divided into 5 regions by the final centerline.

In step 511, the image processing apparatus can detect the erroneous region by comparing curvatures between adjacent regions based on the position of the flash. In other words, in consideration of the flash effect generated in the process of acquiring the image, the image processing apparatus may include a tongue region that is mistaken for the skin region in the image and is deleted, or ii) a tongue region that is mistaken for the tongue region in the image It is possible to perform correction for the area.

In step 512, the image processing apparatus can correct the boundary between the regions by changing the size of the mask according to the mistaken area. In other words, the image processing apparatus can correct a mistaken area for an area whose boundary is blurred due to a flash effect by using a mask.

In step 513, the image processing apparatus can acquire the finally separated snow area through correction for the mistaken area.

6 is a diagram illustrating a result of acquiring a tongue area from an image according to an embodiment.

Referring to FIG. 6, the tongue image has characteristics similar to those of the adjacent tongue and the surrounding color, as compared with other image extraction. In addition, since the tongue image obtained in the mobile environment has no control of the surrounding environment such as brightness and color, and since the flash that is instantaneously launched to compensate for the surrounding brightness value is positioned next to the lens, Since the boundary of the directly received region is ambiguous, it is difficult to extract the accurate tongue region of the image captured in the mobile environment by the conventional tongue region extraction algorithm.

  Therefore, by applying the above-described algorithm, the present invention can extract a more effective tongue area by canceling the problem that the boundary is blurred or shadow is formed between the tongue and other regions due to the flash effect.

   In other words, the tongue extraction algorithm extracts the tongue area from the image in consideration of the flash effect (shadow and light reflection effect), the shooting distance between the camera and the tongue, and the color characteristics of the tongue, And can extract the tongue area more effectively from the image acquired through the mobile having the non-constant shooting environment.

   Ultimately, by applying this technology, users are expected to acquire tongue images in various mobile environments, and thus provide results on the shape of the tongue. In addition, combining this technology with other applications related to U-Healthcare not only contributes to creating customized / advanced medical services such as health management using the tongue area, It is advantageous in terms of commercialization as it is readily available.

The methods according to embodiments of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software.

While the invention has been shown and described with reference to certain preferred embodiments thereof, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined by the equivalents of the claims, as well as the claims.

101: Image processing device
102: video
103: tongue
104: camera
105: Flash

Claims (15)

A method for extracting a tongue area performed by an image processing apparatus,
Controlling a flash disposed on the basis of a position of a camera mounted on the image processing apparatus;
Acquiring an image of a user including a tongue area and a skin area to which a flash effect is applied according to a degree of uniformity of light in a portion illuminated to the user; predicting a position of the flash through header information of the image;
The skin region excluding the tongue region included in the acquired user's image is primarily deleted by using the threshold value of the colorimetric information (CIE a Thresholding), and the skin region is firstly deleted using the threshold of hue thresholding A second step of deleting a skin area for a skin area that has not been deleted in the tongue area from which noise has been removed;
Extracting an active contour for the tongue region from the images of the skin region deleted;
Based on the extracted outline, the vertical length of the tongue area is calculated, and the lingual area corresponding to the end of the tongue is extracted by dividing into 5 equal parts based on the calculated length of the tongue area in the vertical direction. After setting the initial center line of the tongue connecting the center lines of the trisection lines from the tip of the tongue and setting the final center line for the tongue area by drawing a line perpendicular to the initial center line of the set tongue, Dividing the tongue region into left and right portions at regular intervals;
The lengths from the final center line to the left end and the length to the right end are measured for the tongue regions divided at the predetermined intervals and then the measured lengths are compared to determine the left and right symmetry considering the error rate of the length between the left end and the right end. Correcting the tongue area by applying a curvature calculation to the tongue area; And
Extracting a tongue region from the corrected image
The tongue area extracting method comprising: delete The method according to claim 1,
Wherein the deleting comprises:
And performing a preprocessing operation on a skin region deleted from the image by applying median filtering to remove noise in the image from the skin region based on the threshold value of the colorimetric information. delete delete delete The method according to claim 1,
Wherein the correcting comprises:
A tongue region extracting method for correcting a skin region and a tongue region in an image through an erroneous region and an ROI flattening due to a flash effect generated in the process of acquiring an image based on the outline; The method according to claim 1,
Wherein the correcting comprises:
In consideration of the flash effect generated in the process of acquiring the image,
i) a deleted tongue area mistaken for a skin area in the image, or
ii) Skin area that is mistaken for the tongue area in the image and not deleted
The tongue area extraction method comprising: An image processing apparatus comprising:
And controls the flash disposed on the basis of the position of the camera mounted on the image processing apparatus to acquire an image including a tongue region and a skin region to which a flash effect is applied according to the degree of uniformity of light in a portion illuminated to the user An image acquiring unit for predicting a position of a flash through header information of the image;
The skin region excluding the tongue region included in the obtained image of the user is primarily deleted by using the threshold value of the colorimetric information (CIE a Thresholding), and the skin region is firstly deleted using the threshold of hue thresholding A skin area deleting unit that secondarily deletes a skin area with respect to a skin area that has not been deleted in the tongue area from which noise has been removed;
An outline extracting unit for extracting an outline of the tongue area from the images of the skin area deleted;
Based on the extracted outline, the vertical length of the tongue area is calculated, and the lingual area corresponding to the end of the tongue is extracted by dividing into 5 equal parts based on the calculated length of the tongue area in the vertical direction. After setting the initial center line of the tongue connecting the center lines of the trisection lines from the tip of the tongue and setting the final center line for the tongue area by drawing a line perpendicular to the initial center line of the set tongue, A tongue region dividing unit dividing the tongue region into left and right tongue regions at regular intervals;
The lengths from the final center line to the left end and the length to the right end are measured for the tongue regions divided at the predetermined intervals and then the measured lengths are compared to determine the left and right symmetry considering the error rate of the length between the left end and the right end. A tongue area correction unit for correcting the tongue area by applying curvature calculation to the tongue area; And
A tongue area extracting unit for extracting a tongue area from the corrected image,
And the image processing apparatus. delete 10. The method of claim 9,
Wherein the skin area deleting unit comprises:
And applying median filtering to remove noise in the image, wherein the skin region is deleted based on the threshold value of the colorimetric information, to perform a preprocessing operation on the skin region deleted from the image. delete delete delete 10. The method of claim 9,
Wherein the tongue area correction unit comprises:
In consideration of the flash effect generated in the process of acquiring the image,
i) a deleted tongue area mistaken for a skin area in the image, or
ii) Skin area that is mistaken for the tongue area in the image and not deleted
In the image processing apparatus.

Images