Automatic segmentation of thermal images to support breast cancer diagnosis
DOI:
https://doi.org/10.18046/syt.v12i30.1855Keywords:
Infrared thermography, image segmentation, geometric analysis, breast cancerAbstract
A proposal for segmenting thermographic images that can be used as a pre-processing step in the asymmetric analysis of breast cancer. This proposed segmentation of detecting areas of high temperature gradients, from which geometric regions of interest [Region of Interest, ROI] are defined. Hot spots selected as a reference at the start of the identification of the ROI corresponded to those presented under each breast, then, using a contour following on both sides of the body, it was sought to define the coordinates of the vertices that shaped the region of interest. The results show an average success of 67.5% in the segmentation of the breast region from 40 thermograms, which were captured in patients with raised arms or hands on hips while capturing images at a distance of 1 m camera.Downloads
Published
2014-09-30
Issue
Section
Original Research
License
This journal is licensed under the terms of the CC BY 4.0 licence (https://creativecommons.org/licenses/by/4.0/legalcode).
