Publication: A study on hand vein imaging methods
Loading...
Date
2020-07-01
Authors
Ishak, Muhammad Irshad
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The image of the hand vein is collected in near-infrared (NIR), so the level is not as rich as the image of visible light. NIR image has more performance on relative gray concentration and low contrast image. When the infrared light is weak, the vein texture is immersed in the background. In order to improve the illumination of hand vein patterns effectively, hand vein image enhancement thus take part in important key in hand vein recognition. Image enhancement methods can be used to enhance the visual aspect of an image, or to transform an image to form better acceptable to the later processing, such as segmentation and recognition. The proposed system mainly focusses on enhancement stage which is to enhance the vein image in order to produce a clear visible pattern. In this work, a combination of Laplacian sharpening filter and Contrast Limited Adaptive Histogram Equalization (CLAHE) techniques are employed in dorsal hand vein ROI images. Laplacian sharpening filter is used to restore fine detail to an image which has been smoothed to remove noise. It is to give prominence to vein texture. Then Contrast Limited Adaptive Histogram Equalization (CLAHE) is adopted to enhance the contrast of hand vein image. There are all five stage involve in this work. It begins by ROI image collection stage as the input images used in this project. These images are the region of interest extraction from hand vein images. Then follow by generation of truth images stage. At this stage, ground truth images are manually generated based on the original image obtained using paint.net software. Then in the develop enhanced hand vein imaging technique stage, a combination of Laplacian sharpening filter and CLAHE techniques are employed to give prominence to vein texture and enhance the gray level of the ROI images. Then the enhanced images will be analyzed visually and the visualization result of the images will be compared to see the difference. The following section discusses the segmentation stage, local adaptive threshold is employed to segment out the vein patterns from the background image. Finally, correction of the segmented images will be calculated, and the accuracy will be evaluated. The resultant images are then will be compared with the ground truth images. Simulation results show that the vein texture of image is clear, and it is an effective algorithm combination to enhance hand vein image.