Problem statement


The major problem faced by the doctors
today is the difficulty in accessing veins for
intravenous drug delivery. With improper
detection of veins, several problems such as
bruises, rashes, blood clot etc. could occur.

Subcutaneous fat or dark skin color reduces
the visualization
of blood vessels
underneath the skin.

Gaining intravenous access in children can
be difficult.

AccuVein vein finder

How Does It Work?

Hemoglobin in the blood absorbs infrared light, so one laser uses this invisible light to
detect the hemoglobin in the veins. The second uses a visible red light to project the
vein pattern. Because the hemoglobin in the blood absorbs the infrared light anywhere
there are vessels, the light glows back red when it shines on soft tissue but black when
it illuminates a vessel.
In other words, when an injector holds the device above the skin, veins appear
noticeably different from the surrounding tissue. The device looks at the invisible
infrared vein pattern and uses advanced, real-time processing algorithms to detect the
position of the veins under the skin. It then re-projects the vein pattern back on the
skin, forming a crisp digital image. The vasculature clearly shows up on the skin’s
surface, aiding in vein location and avoidance.

i,In NIR , the ring of LEDs are used to get the clear

image of the preferred frame component with infrared light.

This ring constitutes a circuit driver and the power supply

could be provided from the pc the usage of a USB cable

ii. This reflected light rays have been captured by an IR

camera. CCD cameras are chosen for NIR because of its

sensitivity to IR, which gives a better picture quality. One

more advantage is camera being cost effective i.e lesser cost.

The camera really works well on laptop which uses windows

OS. The CCD camera has sufficient spatial resolution to get

clear vein pictures. CCD cameras as shown in fig.5, have

1000nm range of wavelength and the human eye cannot detect

wavelengths which is greater than 750nm.

the CCD camera will be fed into the ROI Component. ROI is

mainly used to select the region of interest that wishes to get .

NIR light source when projected shows the complete images

which has captured using CCD camera , so only ROI is used

to get highlighted segment of an image. Reason to carry out

ROI step as utilizing the whole image takes much time and the

final captured image might not be much accurate. So ROI is

used to acquire the required part of the image.

iv. In Pre-processing , Output of ROI , which have been

captured from IR camera will be fed to the Broad-cam chip

(Raspberry Pi3), which will carry out the pre-processing steps

to get the good quality of images at the end. Pre-processing

needs to be performed before passing images to external

source as shown in fig.6.

a. Required part of image obtained in ROI step undergoes

pre-processing

b. Image processing algorithm has been applied to show

the exact vein pattern of arm/wrist/ limb, to remove noise and

also to enhance the quality of input image that have acquired.

Vascular Imaging Algorithm has shown the process for

getting various patterns of blood vessel to get a images which

are contrast using the near-infrared imaging system.

c. Segmentation using Threshold : Once ROI outputs the

required image part and feeds it to the Broadcom chip, the

Image processing algorithm which are embedded in the chip

will eliminate noise and there by gives the enhanced images.

pattern from the background images. Obtained vein patterns

will be considered and isolated from rest of the images. If this

step not executed properly, noises will be increased. If we

don't get proper images of vein, obviously it will be of no use

again. The concept of Thresholding is used to get converted

the gray scale vein images to the binary images.

d. Filter: The output of segmentation component consists

of gray scale images that might have some unclear and also

unwanted shadows in the images of veins which are displayed.

So it has fed into the filter to upgrade the nature of image

obtained. In order to block the unwanted visible radiations

which are below 720nm allows the near infrared radiation to

fed to the transmission filter . The filter can now effectively

block more than 90% of the visible radiations. The output of

this filter will be the enhanced set of images which is of good

quality

v. In Post-processor, the operation of this component is

performed after the data received from the external source but

before feeding for the display unit. It checks with the

database to know whether the obtained image after the

processing is matching with the template that is stored in

database priory

Focus this
The processor usually consists of Broad-cam chip and the laptop that contains required
software to get effective image patterns. Accuvein device , which is based on NIR , makes use of
OPEN CV (Computer Vision) Library which is supported by the Python Programming
Language to perform the necessary enhancement of the information obtained from the
acquisition system. The Broadcom chip consists of another USB cable interface from the laptop
and an Ethernet cable toget connected as an interface between the Broadcom chip and the laptop.
The images obtained are then undergo processing using the built in language of the
Broadcom chip, which uses Python and is supported by Open CV Image processing library
saved on the SD card and that will be inserted into the broadcom chip along with the boot
software called Raspbian Jessie. The results of this may provide a binary image which
demonstrates the foreground image and the history image based at the contrast greater
picture[10]. Filter takes care of removing the shadow and faint images and gives the enhanced
vein pattern after this processing step. Then post-processing will be applied to check out the
shape match in database, if it is matching it
will be displayed in the display unit with much good presentation of vein patterns.

Methodology
Schematic diagram of the methodology is
shown in Fig.1. The visual, near infrared and
infrared images are captured and processed by
different image processing techniques to
enhance the visualisation of veins. Regions of
interest are selected in all images, processed
by image processing techniques presented in
Fig.1 and prepared for a comparison process.