KR102176196B1 - Real-time detection device and method of superficial vein - Google Patents

Abstract

The present invention detects the shape of superficial veins by analyzing the reaction specificity of hemoglobin in blood in a specific wavelength range in a near-infrared (N-IR) irradiation environment, and applies a high-speed image processing algorithm to reduce blood vessels. It relates to an apparatus and method for real-time detection of superficial veins that are projected on the original skin surface in real time.

Description

Real-time detection device and method of superficial vein}

The present invention detects the shape of superficial veins by analyzing the reaction specificity of hemoglobin in blood in a specific wavelength range in a near-infrared (N-IR) irradiation environment, and applies a high-speed image processing algorithm to reduce blood vessels. It relates to an apparatus and method for real-time detection of superficial veins that are projected on the original skin surface in real time.

IV injection is one of the important medical procedures performed in many medical fields.In most cases, skilled medical personnel perform IV injection without much difficulty, but sometimes obese patients or peculiar constitutions that are difficult to identify blood vessels with the naked eye. In the case of patients and child patients, IV injections are difficult.

In particular, in the case of a patient who is hospitalized for a long time due to a serious disease, it is necessary to administer Ringer's fluid for almost 24 hours, so it must be kept in an IV injection state. In this case, if the blood vessel is weakened by repeated infusion and repeated injections avoiding weak areas, Eventually, the arm, hand, and ankle are no longer able to find a location for IV injection, and eventually, a case of injecting drugs through the insertion of a central venous line occurs.

For this reason, when a drug is injected into a place other than a blood vessel during the IV injection process, symptoms such as infiltration, hematoma, nerve damage, and acute compartment syndrome may occur.

In the above, infiltration refers to the leakage of intravenous fluid to surrounding tissues outside the vein, and hematoma refers to a left injury formed by bleeding to the surrounding tissues at the site of the intravenous injection, and nerve damage [Nerve damage] It occurs when the nerve sheath is actually invaded when the needle is inserted or when the nerve is compressed by an intravenous catheter for a long period of time. Acute compartment syndrome is limited to the pressure in the space confined by trauma. It increases to the range and leads to a decrease in blood flow, which means that tissue ischemia, necrosis, and functional damage occur.

On the other hand, in the case of plastic surgery/dermatology, blood vessels must be avoided during cosmetic procedures in most cases, and in the case of filler (filler) treatment, which is popular in recent years, the procedure must be performed on the face where a large amount of microvessels and nerves are distributed. Accidents due to incorrect procedures tend to occur frequently.

Therefore, there is an urgent need to develop a technology capable of grasping the exact location and state of blood vessels.

Registered Patent 10-1494638

An object of the present invention is to grasp blood vessel information using the reflection principle of near-infrared (N-IR), reconstruct the detected blood vessel information into an image in the visible area, and accurately project the measured in-situ in real time. Through (Projection), IV injection (or filler surgery, etc.) can be performed while grasping the location and shape of blood vessels that cannot be identified with the naked eye, and medical accidents occurring in the IV injection process and plastic surgery It is to provide a real-time superficial vein detection device and method that can dramatically reduce the side effects and reduce the burden on medical personnel, thereby enabling active treatment for patients.

In addition, to provide an apparatus and method for detecting a superficial vein in real time to output an accurate and clear blood vessel image by preventing distortion or noise from occurring during the superficial vein detection process.

A feature of the present invention for achieving the above object is a near-infrared light source module for irradiating the near-infrared ray of 800 ~ 950nm band to the skin; A camera module configured to include at least one lens for obtaining an image of a subcutaneous blood vessel and a near-infrared (IR) image sensor on which an image obtained from the lens is formed to obtain subcutaneous blood vessel data; An image processor module for image processing the subcutaneous blood vessel data transmitted from the camera module; A digital light processing (DLP) module for outputting a blood vessel image output from the image processor module on a photographed skin area in real time; There is a superficial vein real-time detection device comprising a.

In addition, another feature of the present invention is a pre-processing process of correcting optical distortion of a lens and correcting a projection angle error of a DLP using the superficial vein real-time detection device: the pre-processing Geometric transformation process that performs geometric transformation including perspective transformation in the image acquired through the process: Histogram equalization (increase the contrast of the acquired image) by setting a separate Contrast Limiting value for each block to equalize Equalization [CLAHE (Contrast Limited Adaptive Histogram Equalization)] process that performs (Equalization): Noise removal [Median Blur] process that removes increased noise after applying CLAHE: Invert [Invert] process that reverses the contrast of the darkened blood vessel image : Image segmentation and shape detection through binarization [Binarization] process: It is in a method of real-time detection of superficial veins, including.

According to the present invention having the configuration as described above, the blood vessel image is accurately projected in-situ in real time, thereby significantly reducing medical accidents and side effects that occur during the IV injection process, and avoiding the blood vessels to be injected. By not damaging blood vessels when necessary, it can reduce the burden of medical personnel in the situation of IV injections or plastic surgery, enabling active treatment for patients, and increasing confidence in medical staff through accurate procedures even for patients. It works.

Accordingly, it helps to find veins of patients who are difficult to find veins in the blood collection room, such as the elderly and children, to increase work efficiency by reducing the burden of blood collector's mistakes, and to have a sense of stability as patients and guardians can check them with the naked eye. It has the effect of shortening the blood collection time because it can be given, and because even an inexperienced blood collector can easily collect blood.

In addition, during the procedure at the vascular malformation clinic, the location of the lesion can be checked and marked using ultrasound, and the procedure can be performed while directly observing the skin under the skin rather than performing the procedure or performing the procedure while watching the ultrasound, which is more intuitive and increases the operator's convenience. , By easily distinguishing blood vessels and tissues that cannot be distinguished with the naked eye during surgery, it can contribute to increasing the surgical success rate.

In addition, since injection procedures such as intravenous, intramuscular, and subcutaneous injection in plastic surgery can accurately distinguish the subcutaneous layer, side effects of injection can be prevented.

In addition, blood flow pattern analysis may be used as a simple discriminant test at low cost by helping to diagnose peripheral vascular disease, and may help prevent vascular disease.

1 is a view showing an apparatus for real-time superficial vein detection according to the present invention
2 is a diagram showing a blood vessel detection image by LED wavelength
3 is a diagram showing an operation of a visible light filter
4 and 5 are views showing the action of the visible light filter according to the present invention
6 to 9 are views showing an example of driving a user interface for a mobile device
10 to 15 are views showing a method for real-time superficial vein detection according to the present invention
16 to 18 are diagrams showing an embodiment of a method for real-time detection of changes in blood flow

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

As shown in Figure 1, the superficial vein real-time detection apparatus according to the present invention includes a near-infrared light source module 100 that irradiates the skin with near-infrared rays in a band of 800 to 950 nm; A camera module 200 comprising at least one lens 210 for obtaining an image of a subcutaneous blood vessel and a near-infrared (IR) image sensor 220 on which an image obtained from the lens 210 is formed; An image processor module 300 for image processing the subcutaneous blood vessel data transmitted from the camera module 200; And a digital light processing (DLP) module (hereinafter referred to as a DLP module) 400 that outputs a blood vessel image output from the image processor module 300 on a photographed skin area in real time. .

In the drawings, reference numeral 500 denotes a communication module for communication with an external device, and 600 denotes an input/output module such as a menu selection button for power on/off and a connector for charging or data transmission.

In the above, the near-infrared light source module 100 uses an 800-950nm LED element, and in order to check the optimal near-infrared wavelength, a number of commercially available near-infrared LEDs with different wavelengths were used to measure subcutaneous blood vessel images. As shown in Fig. 2, excellent images were obtained in a 850nm LED device (diffusion angle 45°).

Meanwhile, the output of the near-infrared light source module 100 is adjusted and the sensitivity of the near-infrared image sensor is adjusted accordingly to enable detailed image control. For example, if the output of the near-infrared light source module is increased, the clarity of blood vessels can be increased, but Since the detail disappears, microvessels cannot be seen, and if the output of the near-infrared light source module is weakened to see microvessels, there is a correlation that the sharpness of large blood vessels is weakened. Accordingly, the user can adjust the output of the near-infrared light source module and the sensitivity of the near-infrared image sensor according to the intended use.

The lens 210 constituting the camera module 200 is a C-Mount type lens. Non-IR Coating Lenses are used so that the IR band can be sufficiently transmitted through the lens, and even at a short distance. A phenomenon in which the minimum focal length is 15 cm or less so that an image can be formed on the infrared (IR) image sensor 220, and the blood vessel image projected from the DLP module 400 interferes with the IR image sensor 220 (flickering). In order to prevent this, as shown in FIGS. 1 and 3, a visible light filter 230 which is a separate 720 nm low-cut filter is provided on the lens 210.

The visible light filter 230 is coated with a visible light blocking material on the rear surface of the lens, and conventionally, a method of attaching a filter to the front of the lens and the LED light source was used, but in this case, the loss of output LED light due to the scattering effect of the filter itself and Since the resolution of the acquired image is lowered, in the present invention, a clear image can be obtained without loss of LED light intensity and resolution of the acquired image by coating and filtering a visible light blocking material on the rear surface of the lens.

Therefore, when the visible light filter 230 is applied, the light of the ambient light such as sunlight or lamp light, the flickering image, and infinite repetition of the Self Image of Vein are blocked, and the near-infrared image (N- IR Image of Vein) can be passed.

In the above, flickering is a structural problem caused by the DLP module outputting each RGB light source of the image to a rotating mirror, and the infinite repetition of self-image is a problem that occurs while reproducing the image in the original position to be measured, that is, the reproduced image again It is a problem that flows into the measurement process.

As a result of measuring the performance of the visible light filter 230 with an optical spectroscope, it can be seen that optical interference of 720 nm or less is completely blocked as shown in FIG. 4, and FIG. 5 shows before/after using the visible light filter. It shows an image, and it can be seen that the interference disappears after use.

The image sensor 220 is an N-IR CMOS sensor, which is a specialized image sensor having a QE (Quantum Efficiency) characteristic that is about twice as high as a Bayer Color or Monochrome type image sensor in a near-infrared (N-IR) wavelength band. Processor) built-in ROI (Random Programmable Region of Interest) function allows you to cut and use the image area without putting a load on the external processor, and it supports functions such as AEC (Automatic Exposure Control) and HDR (High Dynamic Range) to provide precise blood vessels. By enabling imaging, it is possible to reconstruct an image with a higher resolution in the superficial vein region, and thus it is possible to grasp even microvessels that are not well recognized by conventional vascular detection equipment.

At this time, quantum efficiency (QE) refers to the rate at which photons or electrons of light quantized among materials are converted into photons or electrons of different energy. It refers to the ratio of the photons absorbed by the device and the converted electrons, and the quantum efficiency in the image sensor refers to the visibility and saturation capacity of the image.

In addition, the image sensor 220 changes the HDMI image output to parallel 24-bit, and the camera image input also performs a graphic acceleration function by pin mapping from the MIPI method, which is a conventional serial input method, to a parallel 12-bit gray scale. And change the OS kernel configuration and port remapping of all other pins to prevent collisions.

The image processor module 300 uses an I.MX6Quad processor, which is Freescale Semiconductor's top-level multimedia processor focused on high performance and low power consumption, and is a high-end portable media player with HD-level video processing. HD video capability) and is designed based on the ARM Cortex-A9 MPCore platform.

In addition, the image processor module 300 is configured to include a PMIC for separate battery type power management.

The DLP module 400 is composed of a driver part that adjusts an image projection light source and an engine part that outputs an actual image image.

In addition, the lens tube is manufactured so that the focal length of the projection can be adjusted to be less than 15cm, and the port mapping of the video signal is changed to be implemented on the DLP controller board, and a dedicated device driver is developed to the controller board. Porting.

In the present invention configured as described above, recognition of the camera module 200 by shifting the optical axis of the camera module 200 and the DLP module 400 to match the recognition image of the camera module 200 and the projection area of the DLP module 400 The area and the projection area of the DLP module 400 are matched, and distortion caused by this is corrected by the image processor module 300.

Alternatively, in order to match the recognition image of the camera module and the projection area of the DLP module, an optical mirror is mounted to match the optical axis.

On the other hand, the superficial vein real-time detection device is further provided with a communication module 500 such as Wi-Fi and Bluetooth to transfer the obtained superficial vein image to an external device, that is, a mobile device such as a mobile phone, a person in charge terminal (desktop, laptop, etc.), a server, etc. It transmits and allows the operation to be controlled by a mobile device or a person in charge terminal. An example of a user interface for a mobile device for this will be described below with reference to FIG. 6.

The user interface for the mobile device includes a menu that variously adjusts and outputs the output of the detected superficial vein image, as shown in FIG. 7, and the remaining battery capacity of the superficial vein real-time detection device, enlargement and reduction of an image. It is a streaming menu with detection algorithm applied, and the Fine Detail menu is a streaming menu with higher resolution for precise analysis (blood vessel detection algorithm only applies up to CLAHE stage 2), but not in real time, and the Blook Flow menu sets the resolution to observe the relative change in blood flow. This is an image acquired through cross-lighting of 770nm LED and 850nm LED with a high streaming menu, not a real-time image.

In addition, as shown in FIG. 8, the user interface for the mobile device includes a menu for displaying blood vessels in black or inverting the image to brighten blood vessels, as shown in FIG. Likewise, a color selection (Option: Green (Default)/Blue/Red) menu to be used for the bright part is provided, and an interlocking setting menu for controlling the amount of light of the LED light source and interlocking with the superficial vein detection device is provided.

In addition, by adopting a natural air cooling method, the superficial vein real-time detection device can be used in an operating room, etc., and a blood vessel image processed in high resolution can be streamed in real time to a mobile device such as a tablet, as detailed analysis data of the superficial vein. It can be used for vascular malformation analysis, vascular identification of internal organs in open surgery, and retinal vascular degeneration in ophthalmology.

In addition, the superficial vein image acquired through the superficial vein real-time detection device is transmitted to the server, and by applying an artificial intelligence (AI) learning technique specialized for medical image analysis in the server, it is optimal for IV injection. The location of the puncture (centesis) is recommended, but the optimal puncture location is suggested by weighting through the puncture image obtained during the actual puncture and the expert's puncture location recommendation, and the puncture location proposed from the server is the superficial vein of the present invention in real time. It is transmitted to the detection device and projected onto the skin through the DLP module.

A method for real-time superficial vein detection using the device will be described below.

The superficial vein real-time detection method is a pre-processing process of correcting the optical distortion of the lens and correcting the projection angle error of the DLP: Perspective transformation in the image obtained through the pre-processing process. Geometric transformation process that performs geometric transformation including (Transform): Histogram equalization (contrast limited adaptive histogram (CLAHE (Contrast Limited Adaptive Histogram)) that performs equalization by setting a separate Contrast Limiting value for each block as a method of leveling the histogram Equalization)] Process: Noise removal [Median Blur] process of removing increased noise after applying CLAHE: Inversion [Invert] process of inverting the contrast of the darkly displayed blood vessel image: Shape detection through image segmentation and binarization [Binarization] process: Consists of including.

Optical distortion correction in the pre-processing process will be described.

An image acquired through a lens basically has a distortion phenomenon, and major distortions include radial distortion and tangential distortion.

Among them, radial distortion is caused by the refractive index of the convex lens, and the degree of distortion of the image is determined by the distance from the center (the distance from the center of the image, the more the straight line is distorted into a curve). It is corrected with Equation 1.

Equation 1

Tangential Distortion is a distortion that occurs when the camera lens and the image sensor are not horizontally aligned or the centering of the lens itself is not aligned during the manufacturing (assembly) process of the camera, and some areas of the image may appear closer than expected. Corrected by Equation 2.

Equation 2

Other physical distortion factors that occur are additionally corrected by applying the Brown-Conrady Model.

Equation 3

Projection angle error correction is to prevent distortion of the image due to the lens angle. Before measuring blood vessels, the DLP module first irradiates the position to detect the reference point (or reference line), acquires the coordinates of the reference point from the image data collected through the image sensor, and uses the camera module and the DLP module as reference coordinates. Align and apply to blood vessel detection.

As an example, as shown in FIG. 10, a reference point, that is, coordinates of each of the four corners (indicated by red dots) by photographing a checker board image projected by the DLP module with a camera, is obtained, and the obtained reference The camera module and the DLP module are aligned with coordinates and applied to blood vessel detection.

That is, in order to match the recognition area of the camera module with the projection area of the DLP module, the optical axis of the camera module is shifted by about 16° so that the blood vessel recognition area of the camera module and the projection area of the DLP module coincide. The distortion is then corrected in the image processor module.

After making the recognition area of the camera module and the projection area of the DLP module coincide as described above, a geometric transformation process is performed on the 2D image acquired from the camera module, which occurs because the optical axes of the camera module and the DLP module do not coincide. Perspective transformation is corrected, and the geometric transformation includes perspective, scaling, translation, rotation, and the like, as shown in FIG. 11.

The homogenization process is performed by the histogram equalization method, and histogram equalization increases the contrast of the 2D image, as shown in FIG. 12, thereby better dispersing the intensity of the histogram, and thereby lowering The image area of the contrast ratio becomes higher.

Therefore, general histogram leveling is useful for images with bright or dark backgrounds and foregrounds, especially for overexposed or underexposed images. However, since it is applied over the entire area, the contrast is improved and the image is There is a problem that detailed area information is lost.

Therefore, in order to solve this problem, as shown in FIG. 13, the histogram leveling technique is changed and applied to the modified CLAHE technique, and the CLAHE is repeated a plurality of times as necessary to perform multiple times.

The Contrast Limited Adaptive Histogram Equalization (CLAHE) divides the image into small blocks (tiles) and then performs histogram equalization for each block. However, if there is a lot of noise in the image, Since noise is also amplified together, it is solved by setting a separate contrast limiting value. In addition, artifacts occurring at the boundary of the tile are removed by applying a bilinear interpolation technique.

In order to remove the increased noise after applying CLAHE, as shown in FIG. 14, Median Blur, which is a noise removal process, is performed, which aligns the values of surrounding pixels and their own values according to the size and ) Is an algorithm to select as its own pixel value.

Thereafter, as shown in FIG. 14, an inverting process is performed to reverse the contrast of the image so that dark blood vessels are expressed brightly.

Finally, the shape detection process (Binarization), that is, the shape of the desired part or object is detected by segmenting the image, which scans the entire input image and if the threshold value is larger than the pixel value, the pixel value at the corresponding location is white. (1 or 255), and if the pixel value is less than the threshold, it is set to black and output, and it is projected to the original position (In-Situ) in real time. The image and the actual projected image are shown in FIG. 15.

On the other hand, it is possible to detect changes in blood flow in real time using the superficial vein image detected as described above, and this real-time detection of blood flow changes detects changes in blood flow from an external device by transmitting a superficial blood vessel image to an external device using a communication module. Alternatively, the superficial vein real-time detection device may detect itself.

For real-time detection of such changes in blood flow, a 760-790nm LED element is added to the LED module, and a plurality of 760-790nm LED elements and 800-950nm LED elements are alternately arranged, and each LED element is alternately lit by a camera module. A blood vessel image including concentration information of Hb for each device is acquired, and as shown in FIG. 16 in the obtained blood vessel image for each LED device, a change in the state of blood flow at a plurality of points in the same blood vessel, that is, a relative change in blood flow or an oxidized Hb band The difference in the concentration ratio of the reduced Hb is determined, and when the relative change in the blood flow or the difference in the concentration ratio of each hemoglobin exceeds a reference value (a reference value to be determined to be normal), it is determined as a blood vessel abnormality and output to the display unit.

At this time, the display unit is selected from a display unit of an external device, a display unit provided in a real-time superficial vein detection device, or a DLP module displayed at the original position.

On the other hand, in the measurement of changes in the state of blood flow, the concentrations of Oxide and Deoxide hemoglobin (Hb) in the blood flow detected in real time by alternately lighting 770 nm and 850 nm LEDs are gradated (Scalarize), respectively. This is to grasp the change in the state of blood flow.

And the blood flow can be deduced from the sum of the reduction and oxidation Hb, and the relative degree of change between the two points is derived by summing the difference in the grayscale change of the point to be measured with a value obtained by gradation in a 24-bit system using the RGB888 interface. .

In addition, specific diseases or lesions can be inferred by grasping the difference in the concentration ratio of oxidized and reduced hemoglobin at multiple measurement points. Diseases that can be identified by measuring these concentration ratio changes include arteriovenous fistulas, arteriovenous malformations, venous malformations, capillaries. Vascular arteriovenous malformations, varicose veins, etc.

In the above, arteriovenous fistula is a symptom of an abnormal short circuit between an artery and a vein without passing through an arteriovenous anastomosis, and can be used to determine the area of a short circuit between an artery and a vein by measuring the change in concentration ratio during surgery connecting arteries and veins.

In addition, arteriovenous malformation refers to a congenital defect that is not connected to capillaries due to the entanglement of arteries and veins, and blood flows quickly due to lack of capillaries, and blood is sent to deformed blood vessels. If the amount of change can be measured, it can help to establish a diagnosis and treatment plan.

Varicose veins, as shown in FIG. 17, is a disease in which blood vessels are abnormally expanded/protruded due to reflux of blood because a valve that blocks blood from the leg from falling down again does not function properly. It mainly occurs in middle-aged women. It causes various complications, swelling, itchiness, and spasmodic tissue necrosis, along with common problems. However, there are no external findings before discoloration or protrusion of blood vessels, so initial diagnosis was generally difficult.

In addition, since the dilated blood vessels are not restored, fundamental treatment is difficult, and if the symptoms are severe, the blood vessels must be removed or sclerotherapy must be used. Therefore, a method of detecting them in the early stage is urgently required.

Accordingly, the present invention relates to varicose veins when the difference in concentration of each oxidized Hb and reduced Hb measured at a plurality of points is significantly different from the composition ratio at a specific point at the plurality of points, as shown in FIG. It is determined, and accordingly, it is possible to diagnose early before external findings such as blood vessel protrusion and discoloration occur.

100: near infrared light source module
200: camera module
210: lens
220: Near infrared (IR) image sensor
300: image processor module
400: Digital Light Processing (DLP) module

Claims (10)

Near-infrared light source module 100 for irradiating the near-infrared rays of the 800 ~ 950nm band to the skin;
A camera module 200 comprising at least one lens 210 for obtaining an image of a subcutaneous blood vessel and a near-infrared (IR) image sensor 220 on which an image obtained from the lens 210 is formed;
An image processor module 300 for image processing the subcutaneous blood vessel data transmitted from the camera module 200;
And a digital light processing (DLP) module (hereinafter referred to as a DLP module) 400 that outputs a blood vessel image output from the image processor module 300 on a photographed skin area in real time. ,
Before the blood vessel image is detected, the DLP module irradiates a reference point or reference line to the skin, and in the image acquired from the camera module, the image processor module processes the image based on the reference point or reference line, and the DLP module performs an image based on the reference point or line Superficial vein real-time detection device, characterized in that outputting. The method of claim 1, wherein the lens 210 constituting the camera module 200 uses a C-Mount type non-IR coating lens, and the minimum focal length is 15 cm or less. Superficial vein real-time detection device. The apparatus of claim 1, wherein a visible light filter (230) is provided by coating a visible light blocking material on the rear surface of the lens (210). The method of claim 1, wherein the image sensor 220 uses an N-IR CMOS image sensor, converts the HDMI image output into parallel 24-bit, and the camera image input is pin-mapping in parallel 12-bit gray scale. Pin mapping) superficial vein real-time detection device, characterized in that. delete The real-time superficial vein detection apparatus according to claim 1, wherein the real-time superficial vein detection apparatus further comprises a communication module to transmit the acquired blood vessel image to an external device in real time. The method of claim 1, wherein in order to match the recognition image of the camera module 200 and the projection area of the DLP module 400, the optical axis of the camera module 200 is shifted by about 16° to match the blood vessel recognition area of the camera module. A superficial vein real-time detection device, characterized in that the projection area of the DLP module is matched, and the distortion caused by this is corrected by the image processor module (300). The method of claim 1, wherein the superficial vein image acquired through the superficial vein real-time detection device is transmitted to a server, and by applying an artificial intelligence (AI) learning technique specialized for medical image analysis in the server, IV injection During the procedure, the optimal puncture (centesis) position is recommended, but the optimal puncture position is suggested by weighting through the puncture image acquired during the actual puncture and the expert's puncture position recommendation, and the puncture position suggested from the server is superficial vein. Superficial vein real-time detection device, characterized in that transmitted to the real-time detection device and projected onto the skin through the DLP module. Using the superficial vein real-time detection device according to any one of claims 1, 2, 3, 4, 6, 7 or 8,
Pre-Processing process of correcting the optical distortion of the lens and correcting the projection angle error of the DLP:
Geometric transformation process for performing geometric transformation including perspective transformation in the image acquired through the preprocessing process:
Equalization [CLAHE (Contrast Limited Adaptive Histogram Equalization)] process by setting a separate Contrast Limiting value for each block as a method of leveling histogram (increases the contrast of the acquired image):
Noise removal [Median Blur] process of removing increased noise after applying CLAHE: Inverting [Invert] process of inverting the contrast of the darkened blood vessel image:
[Binarization] process through image segmentation and binarization: Real-time superficial vein detection method, comprising a. The method of claim 9, wherein the near-infrared light source module constituting the superficial vein real-time detection device includes a plurality of 760-790nm LED devices and 800-950nm LED devices, respectively, and alternately lighting them to generate Hb for each LED device as a camera module. A blood vessel image including the concentration information of is obtained, and a change in the state of blood flow at a plurality of points of the same blood vessel in the obtained blood vessel image for each LED element, that is, the relative change in blood flow or the difference in the concentration ratio of oxidized Hb versus reduced Hb is determined. And, when the relative change in blood flow or the difference in the concentration ratio of each hemoglobin deviates from a reference value to be determined as normal, it is determined as a blood vessel abnormality and output to a display unit.

Images