Light treatment device using lesion image analysis, method of detecting lesion position through lesion image analysis for use therein, and computing device-readable recording medium having the same recorded therein

1. A method of detecting a lesion position through lesion image analysis for use in a light treatment device for treating a lesion region, the method comprising:

acquiring an image of a partial region of a skin of a patient by a camera; and

detecting a lesion position by a controller of the light treatment device, the controller configured to:

normalize the acquired image with respect to light based on a predetermined parameter;

set a plurality of blocks of a predetermined size in the normalized image, scan the normalized image by the set blocks, and calculate normal deviation values of pixels in each of the blocks;

create a binary image in which pixels in each block having normal deviation values lower than a predetermined threshold value and pixels in each block having normal deviation values higher than the predetermined threshold value are distinguished from each other so that a mass which is a region including the lesion position clearly appears in the binary image;

extract the mass from the binary image;

set a mass region including the extracted mass in the binary image;

calculate a normal deviation value of pixels in a region of the acquired image or the normalized image corresponding to the set mass region; and

detect a position at which the calculated normal deviation value is a maximum value as the lesion position.

2. A method of detecting a lesion position through lesion image analysis for use in a light treatment device for treating a lesion region, the method comprising:

acquiring an image of a partial region of a skin of a patient by a camera; and

detecting a lesion position by a controller of the light treatment device, the controller configure to:

normalize the acquired image with respect to light based on a predetermined parameter;

set a plurality of blocks of a predetermined size in the normalized image, scan the normalized image by the set blocks, and calculate normal deviation values of pixels in each of the blocks;

create a binary image in which pixels in each block having normal deviation values lower than a predetermined threshold value and pixels in each block having normal deviation values higher than the predetermined threshold value are distinguished from each other so that a mass which is a region including the lesion position clearly appears in the binary image;

extract the mass from the binary image;

set a mass region including the extracted mass in the binary image;

detect a minimum brightness value of each pixel in a region of the acquired image or the normalized image corresponding to the set mass region; and

detect a pixel group having a predetermined range of brightness values based on the detected minimum brightness value as the lesion region.

3. The method according to claim 2 , wherein the step of detecting the pixel group comprises, in a case in which a plurality of pixel groups is detected, detecting one of the pixel groups having a highest deviation with respect to an average of brightness values of all pixels and detecting a position in the detected pixel group at which the brightness value of the pixel is a minimum as the lesion position.

4. The method according to claim 1 , wherein the step of detecting the position comprises,

detecting the lesion position through different processes in a case in which a single position at which the calculated normal deviation value is a maximum value is detected and in a case in which a plurality of positions at each of which the calculated normal deviation is a maximum are detected.

5. The method according to claim 4 , comprising:

in the case in which the plurality of positions are detected, calculating a Euclidean distance between points at all of the detected positions; and

in a case in which the calculated Euclidean distance is less than a predetermined value, selecting one of the detected positions and determining the selected position to be the lesion position.

6. The method according to claim 4 , comprising:

in the case in which the plurality of positions are detected, calculating a Euclidean distance between points at all of the detected positions; and

in a case in which the calculated Euclidean distance is greater than a predetermined value, detecting pixels having normal deviation values higher than a threshold value set in advance for the normal deviation value calculated at the step of calculating the normal deviation value of each pixel; and

calculating the Euclidean distance between the detected pixels to detect a pixel group having pixels gathered in a predetermined value or lower and detecting a position in the detected pixel group at which the normal deviation value is a maximum as the lesion position.

7. The method according to claim 1 , further comprising:

calculating a covariance for positions of the detected pixels and calculating a Mahalanobis distance between points based on the calculated covariance; and

detecting a pixel group having the calculated Mahalanobis distance less than a predetermined value as the lesion position.