KR101667917B1 - Multipurpose skin picturization apparatus using the filter converter - Google Patents

Abstract

The present invention can selectively use a long-wavelength transmission filter (LPF), a short-wavelength transmission filter (SPF) and a band-pass filter (BPF) so that skin image and subcutaneous information (blood vessel / blood flow) To a multi-purpose skin imaging device using a filter conversion device.
The multi-purpose skin imaging apparatus of the present invention comprises an annular light source and includes a laser diode and a light emitting diode (LED) for each of ultraviolet rays, visible rays, infrared rays and near-infrared rays, A light source unit configured to simultaneously light a laser diode and a light emitting diode (LED) with respect to light of one of a light ray, an infrared ray and a near-infrared ray; A long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter. The long wavelength transmission filter, the short wavelength transmission filter, and the band transmission filter are disposed behind the light source unit, A filter conversion unit configured to be positioned in front; And an image acquiring unit that acquires light incident through the filter converting unit, the image acquiring unit including a CCD camera and a lens capable of transmitting visible light to infrared light.

Description

[0001] The present invention relates to a multi-purpose skin picturing apparatus using the filter converter,

The present invention can selectively use a long-wavelength transmission filter (LPF), a short-wavelength transmission filter (SPF), and a band-pass filter (BPF), thereby providing a single skin imaging device, And an image related to subcutaneous information (blood vessel / blood flow).

The medical image to be acquired varies depending on various purposes and depending on the site, such as when the skin is used for cosmetic purposes and when the skin is used for various therapeutic purposes. In particular, the filter (optical filter) used may be different depending on the image to be acquired even in the same region.

In general, an optical filter is an optical element that transmits only light having a predetermined property (for example, light in a specific wavelength range) of the incident light and does not transmit the other light, and is expensive.

A short-wavelength transmit filter (short-pass filter, SPF), which is a filter that transmits light on a short wavelength side and blocks light on a long wavelength side, and a long-wavelength transmit filter (Long pass filter, LPF), and a band pass filter (band pass filter, BPF), which is a filter selectively transmitting only light of a specific wavelength band.

There is a need for a filter conversion device capable of selectively using various filters in order to easily and simply acquire various medical images with one skin imaging device.

Korean Patent No. 10-1260288 entitled " Multifunctional Facial Imaging Device based on a Stereoscopic Adapter for Stereoscopic Imaging " relates to a multifunctional face imaging device based on a stereoscopic adapter for stereoscopic imaging, Dimensional image information with objective diagnostic parameters for 3D-based physiological changes. In Korean Patent No. 10-1260288, a filter unit having a zero-degree polarizing filter, a 45-degree polarizing filter, a 90-degree polarizing filter, and an ultraviolet blocking filter is mounted on a wheel. Generally, a polarizing filter is a filter that utilizes polarizing light, which is a characteristic of light that passes light in a certain direction among lights coming from various directions or blocks light in a certain direction, and is disclosed in Korean Patent No. 10-1260288 (LPF), a short wavelength transmission filter (SPF), and a band pass filter (BPF).

Korean Patent Laid-Open Publication No. 10-2013-0137449 relates to a multi-band filter array-based camera system and image processing method thereof, and is an image processing apparatus having an image sensor including an RGB color filter and a near infrared ray (NIR) The image processing apparatus fuses an RGB color image and a monochrome image (NIR image) to generate a fusion image including both color information and near-infrared information. Therefore, it is possible to output an optimal image by reflecting both the frequency characteristics of the image input through the RGB pixel and the frequency characteristics of the image input through the NIR pixel. In the case of Korean Patent Laid-Open Publication No. 10-2013-0137449, a long wavelength transmission filter (LPF), a short wavelength transmission filter (SPF), and a band pass filter (BPF) can not be selectively used.

Korean Patent No. 10-0952853 relates to a system and an imaging method for obtaining a multi-infrared image, including a zoom lens for magnifying or focusing an image coming out from a tissue, a visible ray region of light passing through the zoom lens, The infrared region includes a transparent color-selection mirror, a visible light camera that receives light in the visible light region reflected through the color-selection mirror, and filters that divide the light in the infrared region transmitted through the color-selection mirror into specific wavelength regions And a infrared camera which receives the light of the infrared region passing through the filter wheel, and can obtain a mixed image in the visible light region and the near-infrared wavelength region of 700-900 nm, Add fluorescence of five different infrared wavelength bands to different mixed colors to obtain multiple images Can. In the case of Korean Patent No. 10-0952853, a long wavelength transmission filter (LPF), a short wavelength transmission filter (SPF), and a band pass filter (BPF) can not be selectively used.

A problem to be solved by the present invention is to provide an apparatus and method for selectively using a long wavelength transmission filter (LPF), a short wavelength transmission filter (SPF), and a band pass filter (BPF) , A skin image, and subcutaneous information (blood vessel / blood flow) -related images.

Another object of the present invention is to provide a multi-purpose skin imaging apparatus using near-infrared rays capable of imaging blood vessel distribution and blood flow simultaneously by using a near infrared ray laser having a wavelength of 800 nm or more, which is penetrable to the depth of blood vessels in the subcutaneous tissue of a human body.

Another problem to be solved by the present invention is to simultaneously obtain subcutaneous blood vessel distribution and blood flow signal by simultaneously irradiating a near-infrared LED (i.e., a near-infrared LED array) and a near-infrared laser, And to provide a multi-purpose skin imaging device, which constitutes a compact optical system.

(단, f_highboost(x,y)는 고주파 지원 필터링된 영상을 나타내며, f(x,y)는 원 영상을 나타내며, A는 0 보다 크고 1 보다 작은 실수임)
에 의해 구하는, 다목적 피부 영상화장치를 제공하는 것이다.Another problem to be solved by the present invention is that a near infrared ray image acquired through a CCD camera is subjected to an image processing process in real time to emphasize a blood vessel pattern, and a laser speckle index (LSI) And a resultant image is outputted in real time using a video output device (monitor or the like) after reconstructing an image so that a part and a slow part can be visually confirmed.
A further object of the present invention is to provide an image processing apparatus and method, in which an image processing unit obtains a high-pass filtered image by performing high-pass filtering on an original image, which is an image received from an image obtaining unit, Frequency supporting filter for sharpening a high-frequency component.
Another object of the present invention is to provide an image processing apparatus,

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)
To provide a multi-purpose skin imaging device.

According to an aspect of the present invention, there is provided a multi-purpose skin imaging apparatus comprising an annular light source and having a light source of ultraviolet rays, visible rays, infrared rays, and near-infrared rays, A light source unit configured to provide one of infrared light and near-infrared light; A long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter. The long wavelength transmission filter, the short wavelength transmission filter, and the band transmission filter are disposed behind the light source unit, A filter conversion unit configured to be positioned in front; And an image acquiring unit that acquires light incident through the filter converting unit, the image acquiring unit including a CCD camera and a lens capable of transmitting visible light to infrared light.

According to another aspect of the present invention, there is provided a multi-purpose skin imaging apparatus comprising an annular light source and having a laser diode and a light emitting diode (LED) for each of ultraviolet rays, visible rays, infrared rays and near- A light source unit configured to simultaneously light a laser diode and a light emitting diode (LED) with respect to light of one of visible light, infrared light, and near-infrared light; A long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter. The long wavelength transmission filter, the short wavelength transmission filter, and the band transmission filter are disposed behind the light source unit, A filter conversion unit configured to be positioned in front; And an image acquiring unit that acquires light incident through the filter converting unit, the image acquiring unit including a CCD camera and a lens capable of transmitting visible light to infrared light.

The multi-purpose skin imaging apparatus of the present invention performs image processing for emphasizing a blood vessel pattern on an image acquired through an image acquisition unit, and applies a laser speckle index (LSI) on the emphasized blood vessel pattern to visually An image processor for emphasizing and reconstructing an image; And a display unit for receiving and outputting the reconstructed image from the image processing unit.

The long-wavelength transmissive filter transmits a wavelength in the near infrared or infrared region of 750 nm or more, and the short-wavelength transmissive filter transmits ultraviolet region of 400 nm or less in wavelength.

The band pass filter includes a first band pass filter, a second band pass filter, and a third band pass filter, the first band pass filter is a filter that transmits a wavelength band having a central axis of 650 nm, And the third band-pass filter is a filter that transmits a wavelength band having a central axis of 475 nm.

The light source of ultraviolet light, visible light, infrared light, and near infrared light may be a laser diode or a light emitting diode (LED), and the image processing unit may be a computer.

The image processing unit acquires the laser speckle from the image received by the image obtaining unit and analyzes the blood flow information.

The image processing unit obtains a high-pass filtered image by high-pass filtering the original image, which is the image received from the image obtaining unit, and performs high-frequency-assisted filtering to sharpen high-frequency components of the original image using the high- I do.

 The image processing unit processes the high frequency supported filtered image

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)

.

The image processing section performs histogram equalization processing on the image subjected to the high frequency support filtering and performs Gaussian smoothing filtering on the image subjected to the histogram equalization processing.

The present invention also relates to an optical transmission system comprising an annular light source configured to selectively provide one of a light source of ultraviolet light, visible light, infrared light and near-infrared light, a long wavelength transmission filter, a short wavelength transmission filter, An image acquisition unit for acquiring light incident through the filter conversion unit, and an image processing unit for performing image processing for emphasizing a blood vessel pattern in an image acquired through the image acquisition unit, Wherein the image processing unit comprises a low pass filtered image acquiring step of acquiring a low pass filtered image by performing low pass filtering on an original image which is an image received from the image obtaining unit, ; A high pass filtered image acquisition step of subtracting the low pass filtered image from the original image to obtain a high pass filtered image; A high frequency support filtering step of sharpening a high frequency component of the original image using the high pass filtered image; A histogram equalization step of performing histogram equalization processing on the image subjected to the high frequency support filtering; And a Gaussian smoothing filtering step of performing Gaussian smoothing filtering on the image subjected to the histogram equalization processing.

In the low-pass filtered image acquisition step, the mask of the low-

이거나,

Lt; / RTI &

Lt; / RTI >

In the high frequency support filtering step, the high frequency support filtered image

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)

.

In the histogram equalization step, the histogram is calculated by calculating the frequency of the lightness value, the cumulative frequency of each lightness value is calculated, the cumulative frequency of each lightness value is normalized, and the cumulative frequency of the normalized lightness value is used as the histogram value. Reorganize.

The Gaussian smoothing filtering step obtains a kernel from a Gaussian function according to a predetermined standard deviation and obtains a Gaussian smoothing filtered image by convoluting the kernel and the image subjected to the histogram equalization processing .

The present invention also relates to an optical transmission system comprising an annular light source configured to selectively provide one of a light source of ultraviolet light, visible light, infrared light and near-infrared light, a long wavelength transmission filter, a short wavelength transmission filter, An image acquisition unit for acquiring light incident through the filter conversion unit, and an image processing unit for performing image processing for emphasizing a blood vessel pattern in an image acquired through the image acquisition unit, The method comprising the steps of: extracting a region of interest using a thresholding technique from an image received from the image acquiring unit and adjusting the size of the extracted region of interest; Extraction step; A Gaussian low-pass filtering step of applying a Gaussian low-pass filter to the image output in the image extracting step; A normalization step of normalizing an image brightness of a pixel in a Gaussian low-pass filtered image in a Gaussian low-pass filtering step to reconstruct an image; The image output from the normalization step is dilated by a dilation technique. The image output from the normalization step is eroded by the erosion technique. The pixel value of the expanded image and the average value of the pixel values of the eroded image are Averaging the expanded pixels and the eroded pixels to reconstruct the image; And binarizing the image by dividing the pixel value of the image output at the averaging step of the expanded pixel and the eroded pixel into a blood vessel and a skin region based on a predetermined threshold value and detecting the binarized image .

The Gaussian low-pass filter is a Gaussian low-pass filter having a size of 5 × 5 and a standard deviation (sigma) value of 0.8.

In the normalization step, the image brightness of a pixel in a Gaussian low-pass filtered image is compared with a reference value m in a Gaussian low-pass filtering step. If the image brightness of the pixel is smaller than the reference value m, ,

And if the image brightness of the pixel is larger than the reference value m,

(Where, I _normalized (x, y) is the image of the normalized brightness of pixels, _original I (x, y) is the image brightness of the pixel prior to normalization, var is distributed responsibility of the image brightness of the pixel in the image)

Lt; / RTI >

The multi-purpose skin imaging apparatus of the present invention can selectively use a long wavelength transmission filter (LPF), a short wavelength transmission filter (SPF), and a band pass filter (BPF) / Blood flow) related images.

(단, f_highboost(x,y)는 고주파 지원 필터링된 영상을 나타내며, f(x,y)는 원 영상을 나타내며, A는 0 보다 크고 1 보다 작은 실수임)
에 의해 구한다.In addition, according to the present invention, it is possible to analyze the blood vessel state of a diseased part in real time, and it can be applied to important medical actions such as simple medical actions such as blood collection, diagnosis and surgery, and can improve accuracy and safety of medical technology.
Further, the present invention provides a multi-purpose skin imaging apparatus using near-infrared rays capable of imaging blood vessel distribution and blood flow simultaneously by using a near-infrared laser having a wavelength of 800 nm or more, which can penetrate to the depth of blood vessels in the subcutaneous tissue of a human body.
In addition, the present invention provides a method for simultaneously obtaining a subcutaneous blood vessel distribution and a blood flow signal by simultaneously irradiating a near-infrared LED (i.e., a near-infrared LED array) and a near-infrared laser, A multi-purpose skin imaging device.
In addition, in the present invention, a near-infrared image obtained through a CCD camera is subjected to an image processing process in real time to emphasize a blood vessel pattern, and a laser speckle index (LSI) is applied on an emphasized blood vessel pattern, And then outputs the resultant image in real time using a video output device (such as a monitor) after reconstructing the video.
According to another aspect of the present invention, there is provided an image processing apparatus including an image processing unit for performing high-pass filtering on an original image, which is an image received from an image obtaining unit, to obtain a high-pass filtered image, Frequency support filtering is performed.
The present invention also relates to a method and apparatus for generating high frequency support filtered images

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)
.

1 is a block diagram schematically showing the configuration of a multi-purpose skin imaging apparatus using the filter conversion apparatus of the present invention.
FIG. 2 is an explanatory diagram schematically illustrating an embodiment of the multi-purpose skin imaging apparatus of FIG. 1;
3 is a view showing the light source unit 110 of FIG.
4 is a diagram showing the filter conversion unit 130 of FIG.
5 shows the wavelength transmission characteristics of the long wavelength transmission filter.
6 shows the wavelength transmission characteristics of the short wavelength transmission filter.
FIG. 7 is a graph for explaining the characteristics of the first band-pass filter R, the second band-pass filter G and the third band-pass filter B. FIG.
8 is an example of a GUI screen of the multi-purpose skin imaging apparatus of the present invention.
9 is a flowchart illustrating a blood vessel detection algorithm according to an embodiment of the image processing unit of FIG.
10 is a flowchart for explaining a blood vessel detection algorithm according to another embodiment of the image processing unit of FIG.
11 is an example of a 5 × 5 size Gaussian mask.
12 is an explanatory diagram for explaining a process of performing a gray scale expansion operation.
13 is an explanatory view for explaining a procedure of performing a gray scale erosion operation.

Hereinafter, a multi-purpose skin imaging apparatus using the filter conversion apparatus of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically showing the configuration of a multi-purpose skin imaging apparatus using the filter conversion apparatus of the present invention.

The multi-purpose skin imaging apparatus of the present invention includes a light source unit 110, a filter conversion unit 130, an image acquisition unit 150, an image processing unit 200, and a display unit 250.

The light source unit 110 is formed of an annular light source and can selectively provide ultraviolet (UV) visible light (VIS) infrared (IR) light. That is, predetermined illumination light is provided to the skin region desired to be imaged. The LED (for example, a near-infrared LED array) and a laser of a predetermined wavelength (for example, a near-infrared laser) can be simultaneously irradiated from the annular light source 115. By doing so, it is possible to acquire laser speckle and analyze blood flow information.

Laser Speckle Contrast Image (LSCI) is an irregular pattern generated by the interference of light reflected by a laser-irradiated object. Laser Speckle Contrast Image (LSCI) Lt; / RTI >

The filter conversion unit 130 is disposed behind the light source unit 110 and includes a long wavelength transmit filter (LPF), a short wavelength transmit filter (SPF), and a band pass filter (BPF).

The image acquiring unit 150 includes a CCD camera 155 and a lens (not shown) capable of transmitting visible light through infrared light, as means for acquiring light reflected from the skin through a predetermined filter of the filter converting unit 130 .

The image processing unit 200 includes a computer and outputs image information obtained through image processing to an image generated by the image obtaining unit 150 to the display unit 250. That is, the image acquired through the image acquisition unit 150 is subjected to an image processing process in real time to emphasize a predetermined pattern, for example, a blood vessel pattern, and a laser speckle index (LSI) It is possible to reconstruct the image so that the fast part and the slow part can be visually confirmed and output the result in real time using the display part 250. [

The image processing unit 200 enables the user to control the CCD camera 155 in real time through the GUI and processes the image obtained through the CCD camera 155 in real time and outputs the processed image to the display unit 250 .

FIG. 2 is a schematic view illustrating an embodiment of the multi-purpose skin imaging apparatus of FIG. 1, FIG. 3 is a view illustrating the light source unit 110 of FIG. 2, Fig.

The multipurpose skin imaging apparatus includes a CCD camera 155, a rotatable filter conversion apparatus 135 and an annular light source 115 as shown in Fig. 2, in which an annular light source 115 irradiates a predetermined light onto the skin, The light reflected from the skin is incident on the converging lens of the CCD camera 155 through a predetermined filter of the rotatable filter conversion device 135.

3, the annular light source 115 includes a plurality of light sources (LEDs or laser diodes) of ultraviolet (UV), visible light (VIS), infrared (IR) And the light source selected by selecting one of ultraviolet (UV), visible light (VIS), infrared (IR), and near infrared (NIR) lights is turned on in the key operation unit 120. That is, the annular light source 115 may have a total of about 100 UV, VIS, and NIR LEDs arranged in an annular shape, and a predetermined light source can be selected through a switch. The user may select one of ultraviolet (UV) light, visible light (VIS), infrared light (IR), and near infrared light (NIR) through a graphic user interface (GUI) instead of the key operation unit 120.

For example, if a near infrared ray LED (i.e., a near-infrared LED array) and a near-infrared laser are simultaneously irradiated from the annular light source 115 and a subcutaneous vein distribution and a blood flow signal are acquired simultaneously using a near- A compact optical system can be constructed simply without a filter. In this way, the near-infrared image acquired through the CCD camera is subjected to real-time image processing to emphasize the blood vessel pattern. By applying the laser speckle index (LSI) on the emphasized blood vessel pattern, And then outputs the resultant image in real time using the display unit 250. [0053] FIG.

A light source can be added through the socket S provided in the light source housing part 113. [ A diffusion lens (not shown) may be provided at the front end of the annular light source 115. That is, since the diffusion lens and the laser diode can be coupled through the socket S in the annular light source 115, an additional laser light source can be secured.

The rotatable filter inverter 135 is positioned behind the annular light source 115 and includes a long wavelength transmit filter (LPF), a short wavelength transmit filter (SPF), and three band pass filters (BPF) , And can be selectively used.

The rotating type filter conversion device 135 is configured to be rotatable about the center in the shape of a disk, and a desired filter may be positioned in front of the converging lens of the CCD camera 155.

The long wavelength transmission filter (LPF) transmits near infrared rays or infrared rays of more than a wavelength of 750 nm, and FIG. 5 shows a wavelength transmission characteristic of the long wavelength transmission filter.

The near infrared rays transmitted through the long wavelength transmission filter (LPF) are mostly absorbed in the blood vessels and are suitable for simply acquiring and imaging signals of blood vessels. That is, when the near infrared LED is turned on in the annular light source 115 by selecting near infrared rays in the key operation unit 120 and the long wavelength transmit filter (LPF) is selected in the rotary filter conversion apparatus 135, Do. In addition, laser speckles can be obtained by irradiating laser of a similar wavelength, and blood flow information analysis is possible.

Since near infrared rays have a low absorption rate in skin tissue, they can penetrate relatively more deeply than light of other wavelengths and have a high absorption rate in blood vessels. Therefore, it is possible to acquire vascular distribution images through analysis of contrast of images.

The short-wavelength transmission filter (SPF) transmits an ultraviolet ray region having a wavelength of 400 nm or less, and FIG. 6 shows a wavelength transmission characteristic of the filter. Using a short wavelength transmittance filter (SPF), fluorescence images of an ultraviolet region of 400 nm can be obtained to confirm the presence or absence of fluorescent substances in the skin tissue.

(BPF), i.e., a first band pass filter (R), a second band pass filter (R), and a second band pass filter (BPF) for allowing the R, G, and B channels to be selected among visible light ranges of 400 to 750 nm as the band pass filter 7 shows the characteristics of the first band-pass filter R, the second band-pass filter G and the third band-pass filter B, FIG.

The first band pass filter represents an R channel, and is a filter that transmits a wavelength band having a central axis of 650 nm.

The second band pass filter represents a G channel and is a filter that transmits a wavelength band having a central axis of 510 nm.

The third band pass filter represents a B channel, and is a filter that transmits a wavelength band having a central axis of 475 nm.

It is possible to select a wavelength region to be acquired by rotating the rotating type filter conversion apparatus 135. [

FIG. 8 shows an example of a GUI screen of the multi-purpose skin imaging apparatus of the present invention, which includes an image output unit 251, an operation control unit 252, a mode setting and image processing parameter adjustment unit 523.

The user sets the color mode (RGB mode) or the monochrome mode (Gray mode) in the mode setting and image processing parameter adjusting unit 523, sets the filter to be used, and sets the mask size or standard deviation (б, Sigma) is set, and a threshold value is set.

When the user clicks the start button on the operation control unit 252 to execute the image processing, the image is outputted to the image output unit 251. When the image processing is to be terminated, the user clicks the Exit button and terminates , Save the image with the set name.

The image obtained through the CCD camera 155 is first streamed in real time through the image output unit 251 and the user adjusts the operation mode by setting the operation control unit 252 and the mode setting and image processing parameter adjusting unit 523. [ And outputs the newly reconstructed image to the image output unit 251 after changing the output method in the image processing unit 200 or through image processing.

The live streaming of the image of the CCD camera 155 is possible through the start button and the desired frame can be stored as a photo file through the save button. In some cases, the start button is a toggle switch that can be used as a start and stop button. As soon as the stream is started, the Start button changes to the Stop button, and the Stop and Exit buttons are used to stop and exit the program. The image output mode has two modes: a color mode (RGB mode) for outputting an RGB color image using a visible light CCD and a monochrome mode (Gray mode) for outputting a monochrome image for a near-infrared image. A blood vessel detection algorithm is applicable. Some parameters applied to the algorithm can be adjusted in real time in the GUI.

9 is a flowchart illustrating a blood vessel detection algorithm according to an embodiment of the image processing unit of FIG.

In the low pass filtered image acquisition step, the image processing unit 200 low-pass filters an original image (i.e., an original image) received from the image acquisition unit 150, An image is obtained (S110).

In general, low pass filtered image is obtained by performing the steps equation (1), original image (original image) the low-pass filter mask (W _lowpass) and the convolution of the (convolution).

Here, W _lowpass (x, y) represents a low pass filter mask, f _{low pass} (x, y) represents a low pass filtered image, and f _original Image, that is, an original image.

이거나,

일 수 있다.In the present invention, the mask of the low-pass filter

Lt; / RTI &

Lt; / RTI >

In the high pass filtered image acquisition step, a high pass filtered image is obtained by subtracting the low pass filtered image from the original image (S120). That is, the high-pass filtered image can be expressed by Equation (2).

Here, f _highpass (x, y) represents a high-pass filtered image.

In the high frequency support filtering step, a high-boost filter is performed in order to enhance and sharpen the high frequency component of the original image (S130). That is, the original image is convoluted with the coefficient A, and high pass filtered images are added to the original image.

Generally, a high-boost filter can be performed as Equation (3).

Where f _highboost (x, y) represents the high frequency supported filtered image, f (x, y) represents the original image, and A can be a real number greater than 1.5 and less than 1.8.

The high-boost filter of Equation (3) can be expressed as Equation (4).

However, A may be a real number greater than zero and less than one.

In the histogram equalization step, a histogram equalization process is performed on the image subjected to the high-boost filter in the high-frequency support filtering step (S140).

In the histogram of the digital image, the horizontal axis represents the brightness value (brightness value) representing the brightness of the pixel, the vertical axis represents the frequency of the brightness value, and the histogram equalization (smoothing) is dark. By adjusting the histogram of the captured image, It makes it uniform. Generally, in the histogram equalization (smoothing), histograms are calculated by calculating frequency of lightness values, cumulative frequencies are calculated for each lightness value, a minimum cumulative frequency value and a maximum cumulative frequency value among the cumulative frequencies of each lightness value are obtained, The cumulative frequency of each light and dark value is normalized to the difference between the minimum cumulative value and the maximum cumulative value (in some cases, the total number of pixels of the image), and the cumulative frequency of the normalized light and dark values is converted to the brightness value (that is, the histogram value) To reconstruct the image. The histogram equalization (smoothing) method is known from Korean Patent No. 10-0393071, and a detailed description thereof will be omitted.

In the Gaussian smoothing filtering step, a histogram equalization process is performed in the histogram equalization step, and a Gaussian smoothing filtering is performed on the image (S150). That is, the image subjected to the histogram equalization processing is convoluted with Gaussian kernels having different standard deviations to obtain a Gaussian smoothing filtered image. The Gaussian function (G (x, y)) is as follows.

Here, б is the standard deviation.

Given a standard deviation in the Gaussian function, a kernel is obtained, and a Gaussian smoothing filtered (blooded) image is obtained by convoluting the kernel with the image. The Gaussian smoothing filter is also called a Gaussian filter, or Gaussian blur.

10 is a flowchart for explaining a blood vessel detection algorithm according to another embodiment of the image processing unit of FIG.

In the image extracting step, the image processing unit 200 receives an image from the image obtaining unit 150 and extracts an image of a region of interest from the received image (S210). The extraction of the region of interest can separate the background and the region of interest using a thresholding technique in a grayscale image. For example, when the image detected by the image obtaining unit 150 is a hand image, a shape of a hand is detected using a thresholding technique, and only a hand image of interest is cut out .

Here, the image detected by the image obtaining unit 150 may be an image of a hand, and may be a target of detecting a hand, a wrist, or a blood vessel as a region of interest. The image detected by the image acquisition unit 150 of the present invention is not limited to the image of the hand, and the image of the part requiring the blood vessel detection such as the back of the hand, the wrist, the thigh, and the neck can be applied.

In the image size adjusting step, the size of the image of the ROI extracted in the image extracting step is adjusted to a predetermined size (S220).

In the Gaussian low-pass filtering step, a 5 × 5 size and a standard deviation (sigma) value of 0.8 are set in order to remove the speckle noise caused by the light source from the image (ie, the image of the region of interest) A Gaussian low pass filter is applied (S230). That is, a Gaussian low-pass filter is applied to remove noise of point components in an image to obtain a smooth image. Here, a 5 × 5 size Gaussian mask having a standard deviation sigma value of 0.8, that is, a kernel is shown in FIG.

In the normalizing step, the image brightness of a pixel in a Gaussian low-pass filtered image is compared with a gray level reference value m, and if the image brightness of the pixel is smaller than the reference value m, If the image brightness of the pixel is greater than the reference value m, the image brightness of the pixel is normalized using Equation (6) to reconstruct the image (S240).

In Equation 5 and Equation 6, I _normalized (x, y) is the image brightness of the normalized pixel, I _original (x, y) is the image brightness of the pixel before normalization, var is the image brightness And m is the gray level reference value of the image brightness of the pixel, which may be an average of the image brightness of the pixel in the image.

The brightness range redefined by the equations (5) and (6) is easy to detect the blood vessel because the contrast is extended based on the reference value (m).

In other words, in the Gaussian low-pass filtered image, a normalization process is performed to improve distorted image information due to system noise or external negative influence. That is, after the Gaussian low-pass filtering step, normalization is performed so as to improve still image noise or information distorted by the imaging environment, so that dark portions (portions expected to be blood vessels) become darker, Is expected to be brighter.

The image reconstructed through the normalizing step is expanded through a gray scale-based dilation technique (S250) by an average operation step of the diluted pixels and the erosion pixels (S250) the image reconstructed through the normalize step is contracted through a grayscale-based erosion technique (S260), and the pixel value of 0.5 times the pixel value of the expanded image and the pixel value of the eroded (contracted) image (In S270), the pixel values of the expanded image and the average value of the eroded (shrunk) image are reconstructed. That is, after performing the expansion operation and the erosion operation, which are one of the grayscale morphology techniques, an average of the pixel values of the two images is performed to distinguish the line regions (blood vessels). In other words, the reconstructed image through the normalize step approaches the blood vessel pattern from the morphological viewpoint through the gray scale-based dilation and erosion technique.

In the expansion step (S250), the expansion can be obtained by a model called a structuring element as shown in FIG. 12 in an operation of growing or thickening an object. Here, the structural element correspondence values are the values of the upper, lower, left, right, and center around the center value. 12 (a), the corresponding values of the structural elements are {0,67,124,138} when '138' is the center, and the maximum value is '138' '. In FIG. 12C, when '124' is the center, corresponding values of the structural elements {0,109,124,138,191} are outputted, and the maximum value '191' is outputted at the center as shown in FIG. 12B. That is, as shown in FIGS. 12 (a) and 12 (c), the maximum value at the portion where the structural element overlaps the original image is derived, and the shape of the original image is maintained As the size expands, the brightness also increases.

In the erosion step S260, as shown in FIG. 13A, the corresponding values {0,67,109,111,138} of the structural element when '67' is the center, the minimum value is '0' As shown in FIG. 13 (c), the values corresponding to the structural elements when the '109' is the center are 41, 64, 67, 124, and 109, the minimum value '41' is outputted at the center as in d). 13 (a) and 13 (c), the minimum value of the pixel values in the corresponding range is derived while moving the structural element, The width decreases,

Assuming that the structural element is E and each domain is D _I , D _E , the expansion operation expression is defined as shown in Equation (7), and the erosion operation is defined as in Equation (8).

In Equations (7) and (8), D _I is the domain before the expansion or erosion operation, D _E is the domain after the expansion or erosion operation, and I _original is the pixel value.

 In the binarization step of the image, a pixel value of the expanded image and an average of the pixel value of the eroded (contracted) image are performed in the average operation step of the expanded pixel and the eroded pixel, and the pixel value of the reconstructed image is set Based on the threshold value, the blood vessel and the skin region are divided in half (S280), and the binarized image is detected (S290). That is, the derived expansion and erosion images are derived from a threshold image composed of the average value of the brightness of two images, and then the final image is obtained through the binarization process. ThresholdBinary (Threshold = 3) of the Emgucv library can be used to derive the binarized image, and the threshold value can be changed in real time.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all equivalents or equivalent variations thereof are included in the scope of the present invention.

110: light source part 113: light source housing part
115: annular light source 120: key operation unit
130: filter converting unit 135: rotating type filter converting device
150: Image acquisition unit 155: CCD camera
200: image processing unit 250:

Claims (21)

A light source unit comprising an annular light source and provided with a light source of ultraviolet rays, visible rays, infrared rays and near-infrared rays and providing one of ultraviolet rays, visible rays, infrared rays and near-infrared rays according to a control signal received from a key operation unit;
A long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter. The long wavelength transmission filter, the short wavelength transmission filter, and the band transmission filter are disposed behind the light source unit, A filter conversion unit configured to be positioned in front;
An image acquiring unit that acquires light incident through the filter converting unit, the image acquiring unit including a CCD camera and a lens capable of transmitting visible light to infrared light;
A method of performing an image processing for emphasizing a blood vessel pattern on an image acquired through an image acquisition unit and applying a laser speckle index (LSI) on an emphasized blood vessel pattern to visually emphasize a blood flow portion and a slow portion to reconstruct an image A processor;
And,
The image processing unit obtains a high-pass filtered image by high-pass filtering the original image, which is the image received from the image obtaining unit, and performs high-frequency-assisted filtering to sharpen high-frequency components of the original image using the high- Gt; a < / RTI > multi-purpose skin imaging device. And a light emitting diode (LED) for each of ultraviolet rays, visible rays, infrared rays, and near-infrared rays, and is configured to emit one of ultraviolet rays, visible rays, infrared rays, and near-infrared rays in accordance with a control signal received from the key operation unit. A light source unit configured to simultaneously light a laser diode and a light emitting diode (LED) for the light source;
A long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter. The long wavelength transmission filter, the short wavelength transmission filter, and the band transmission filter are disposed behind the light source unit, A filter conversion unit configured to be positioned in front;
An image acquiring unit that acquires light incident through the filter converting unit, the image acquiring unit including a CCD camera and a lens capable of transmitting visible light to infrared light;
A method of performing an image processing for emphasizing a blood vessel pattern on an image acquired through an image acquisition unit and applying a laser speckle index (LSI) on an emphasized blood vessel pattern to visually emphasize a blood flow portion and a slow portion to reconstruct an image A processor;
And,
The image processing unit obtains a high-pass filtered image by high-pass filtering the original image, which is the image received from the image obtaining unit, and performs high-frequency-assisted filtering to sharpen high-frequency components of the original image using the high- Gt; a < / RTI > multi-purpose skin imaging device. delete 3. The method according to any one of claims 1 to 3,
A display unit for receiving the reconstructed image from the image processing unit and outputting the reconstructed image;
Further comprising: an imaging device for imaging the multi-purpose skin. 3. The method according to any one of claims 1 to 3,
Wherein the long wavelength transmittance filter transmits a wavelength in a near infrared to infrared wavelength range of 750 nm or more. 3. The method according to any one of claims 1 to 3,
Characterized in that the short-wavelength transmissive filter transmits an ultraviolet region of less than or equal to 400 nm wavelength. 3. The method according to any one of claims 1 to 3,
The band pass filter includes a first band pass filter, a second band pass filter, and a third band pass filter,
The first band-pass filter is a filter that transmits a wavelength band having a central axis of 650 nm,
The second band-pass filter is a filter that transmits a wavelength band having a central axis of 510 nm,
And the third band-pass filter is a filter that transmits a wavelength band having a central axis of 475 nm. The method according to claim 1,
Characterized in that the light source of ultraviolet light, visible light, infrared light and near infrared light comprises a laser diode or a light emitting diode (LED). delete delete delete The image processing apparatus according to any one of claims 1 to 3,
High frequency support Filtered image

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)
Wherein said at least one skin imaging device is obtained by at least one of the following methods. 13. The image processing apparatus according to claim 12, wherein the image processing unit
Wherein the histogram equalization processing is performed on the image subjected to the high frequency support filtering and the image subjected to the histogram equalization processing is subjected to the Gaussian smoothing filtering. An annular light source configured to selectively provide one of a light source of ultraviolet light, visible light, infrared light, and near-infrared light, one of a long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter located behind the annular light source, An image acquiring unit that acquires light incident through the filter converting unit, and an image processing unit that performs image processing that emphasizes a blood vessel pattern in the image acquired through the image acquiring unit. In the image processing method,
The image processing unit may include a low pass filtered image acquiring step of performing low pass filtering on the original image, which is the image received from the image obtaining unit, to obtain a low pass filtered image;
A high pass filtered image acquisition step of subtracting the low pass filtered image from the original image to obtain a high pass filtered image;
A high frequency support filtering step of sharpening a high frequency component of the original image using the high pass filtered image;
A histogram equalization step of performing histogram equalization processing on the image subjected to the high frequency support filtering;
A Gaussian smoothing filtering step of performing Gaussian smoothing filtering on the image subjected to the histogram equalization processing;
And displaying the image on the screen. 15. The method of claim 14,
In the low-pass filtered image acquisition step, the mask of the low-

Lt; / RTI &

Wherein the image processing device is a multi-purpose skin imaging device. 15. The method of claim 14,
In the high frequency support filtering step, the high frequency support filtered image

(However, _highboost f (x, y) indicates a high-frequency support filtered image, f (x, y) denotes the original image, A is a real number greater than 1 being small and large than 0)
Wherein the image processing method comprises the steps of: 17. The method of claim 16, wherein the histogram equalization step comprises:
The histogram is calculated by calculating the frequency of the intensity value, the cumulative frequency of each brightness value is calculated, the cumulative frequency of each brightness value is normalized, and the image is reconstructed using the cumulative frequency of the normalized brightness value as the histogram value Wherein the image processing method of the multi-purpose skin imaging apparatus comprises: 18. The method of claim 17, wherein the Gaussian smoothing filtering comprises:
Wherein a kernel is obtained from a Gaussian function in accordance with a predetermined standard deviation and a Gaussian smoothing filtered image is obtained by convoluting the image subjected to the histogram equalization processing with the kernel, Image processing method of imaging device. An annular light source configured to selectively provide one of a light source of ultraviolet light, visible light, infrared light, and near-infrared light, one of a long wavelength transmission filter, a short wavelength transmission filter, and a band transmission filter located behind the annular light source, An image acquiring unit that acquires light incident through the filter converting unit, and an image processing unit that performs image processing that emphasizes a blood vessel pattern in the image acquired through the image acquiring unit. In the image processing method,
The image processing unit extracts a region of interest using a thresholding technique from the image received from the image obtaining unit, and adjusts the size of the extracted region of interest.
A Gaussian low-pass filtering step of applying a Gaussian low-pass filter to the image output in the image extracting step;
A normalization step of normalizing an image brightness of a pixel in a Gaussian low-pass filtered image in a Gaussian low-pass filtering step to reconstruct an image;
The image output from the normalization step is dilated by a dilation technique. The image output from the normalization step is eroded by the erosion technique. The pixel value of the expanded image and the average value of the pixel values of the eroded image are Averaging the expanded pixels and the eroded pixels to reconstruct the image;
A binarization step of binarizing the image by dividing the pixel value of the image output from the average computation step of the expanded pixel and the eroded pixel into a blood vessel and a skin area based on a predetermined threshold value;
Wherein the multi-purpose skin imaging device comprises: 20. The method of claim 19,
Wherein the Gaussian low-pass filter is a Gaussian low-pass filter having a size of 5x5 and a standard deviation sigma of 0.8. 20. The method of claim 19, wherein in the normalization step,
In the Gaussian low-pass filtering step, the image brightness of a pixel in a Gaussian low-pass filtered image is compared with a reference value m. If the image brightness of the pixel is smaller than the reference value m,

And if the image brightness of the pixel is larger than the reference value m,

(Where, I _normalized (x, y) is the image of the normalized brightness of pixels, _original I (x, y) is the image brightness of the pixel prior to normalization, var is distributed responsibility of the image brightness of the pixel in the image)
Wherein the image processing unit normalizes the image data by the image processing unit.

Images