Journal
JOURNAL OF MEDICAL SYSTEMS
Volume 39, Issue 4, Pages -Publisher
SPRINGER
DOI: 10.1007/s10916-015-0227-1
Keywords
Health status detection; Non-invasive; Texture feature; Gabor filter; Support vector machine
Funding
- University of Macau [SRG2013-00048-FST]
- Science and Technology Development Fund (FDCT) of Macao SAR [FDCT/128/2013/A]
Ask authors/readers for more resources
Blood tests allow doctors to check for certain diseases and conditions. However, using a syringe to extract the blood can be deemed invasive, slightly painful, and its analysis time consuming. In this paper, we propose a new non-invasive system to detect the health status (Healthy or Diseased) of an individual based on facial block texture features extracted using the Gabor filter. Our system first uses a non-invasive capture device to collect facial images. Next, four facial blocks are located on these images to represent them. Afterwards, each facial block is convolved with a Gabor filter bank to calculate its texture value. Classification is finally performed using K-Nearest Neighbor and Support Vector Machines via a Library for Support Vector Machines (with four kernel functions). The system was tested on a dataset consisting of 100 Healthy and 100 Diseased (with 13 forms of illnesses) samples. Experimental results show that the proposed system can detect the health status with an accuracy of 93 %, a sensitivity of 94 %, a specificity of 92 %, using a combination of the Gabor filters and facial blocks.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available