Loading...
Automatic stitching of medical images using feature based approach
Author(s)
Win, Kyi Pyar
Kitjaidure, Yuttana
Hamamoto, Kazuhiko
Date Issued
January 1, 2019
Type
Article
Abstract
Image stitching is a process of creating a panoramic image by combining multiple images that have overlapping regions of the same scene. It is a challenging topic in image processing, multimedia, and medical applications. The proposed system can be applied in medical applications for scoliosis operations and other long limb operations. The problem of the traditional x-ray machine is the narrow view. So, it can’t produce the large view of body x-ray image in a single frame. Therefore, this problem is solved by combining two or more x-ray images into a panorama one. This paper proposes a system which automatically stitches the x-ray images and produced a panorama x-ray image. The proposed stitching method is based on feature based approach, ORB (Oriented FAST and Rotated BRIEF). In feature detection stage, Oriented FAST approach is used. In feature description stage, Rotated BRIEF approach is applied. The two important criterias for determining the stitching performance are stitched image quality and processing time. Therefore, the aim of my proposed system is to produce a panorama x-ray image with high resolution and low processing time based on feature extraction approach. We compared our proposed method with three different features detectors. SIFT, SURF, and Harris corner detectors were tested and measured the rate of correct features detection and computation time. Finally, we measured the quality of result images that produced by stitching system of different feature based methods. According to the experimental results, ORB approach can produce high quality panorama image with least processing time.
Citation
Advances in Science Technology and Engineering Systems, 4(2), 127-133, 2019