Please use this identifier to cite or link to this item:
|Title:||Image segmentation and colour analysis for wound diagnostics||Authors:||Tay, Ivie Rosabel Ruo Pei||Keywords:||DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision||Issue Date:||2019||Abstract:||In this report, monitoring of a wound through its healing stages is investigated using image processing techniques. It can cause a problem in the medical field because assessment of wounds via visual inspection can be inaccurate due to subjective bias . Therefore, the aim is to separate the wound from the healthy tissue surrounding it using image segmentation and then to analyse its colour as it heals. The effectiveness of several standard image segmentation techniques: Sobel Edge Detection, Canny Edge Detection, Adaptive Thresholding, K-Means, and Fuzzy C-Means, on a set of wound images is studied and their various merits and drawbacks are discussed. Through this investigation it is discovered that although K-Means is a simpler algorithm when compared to the others, it consistently provides the best segmentation for wounds of various kinds. It is also seen that isolation of the wound from image becomes progressively difficult as the wound heals and its texture and colour approaches that of the surrounding healthy skin. Histogram colour analysis on the prominent wound segments obtained using both K-Means and Fuzzy C-Means is carried out. Colour analysis of the wound segments of interest helps to monitor the wound health over a period. Extensive simulation results are shown for various types of wound images both for wound segmentation and colour analysis.||URI:||http://hdl.handle.net/10356/78026||Rights:||Nanyang Technological University||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Updated on Nov 29, 2020
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.