Reliable Biometric Estimation from Colorimetric Data.
Zheng, Ce.
2019
-
Imaging processing for
the color analysis has been explored extensively because of the abundant information
that color can provide. In some specific biomedical patches, colors of certain areas
would change based on the various chemical environment such as pH level, lactate level,
ammonium level, and temperature. By building regression model for color information,
patch's surrounding chemical ... read moreenvironment and temperature can be monitored continuously
through the image of the patch. In this thesis, all desired color areas were segmented
first by applying masks to remove the background and then extract color information of
the single-color area. Considering images are taken under the different light condition,
colors we collected from the image may not be the real color we desired. Calibration is
required to compensate for variations in color due to ambient light. An algorithm using
optimal transport theory has been proposed to solve this light calibration issue. Next,
color information was analyzed to find a relation to the corresponding chemical
information. several regression models were built to predict chemical information
according to colors. Experimental and body trial results prove that our prediction of
chemical environment and temperature by imaging processing achieves the accuracy
requirement.
Thesis (M.S.)--Tufts University, 2019.
Submitted to the Dept. of Electrical Engineering.
Advisor: Shuchin Aeron.
Committee: Shuchin Aeron, Eric Miller, and Liping Liu.
Keyword: Electrical engineering.read less - ID:
- qn59qh39q
- To Cite:
- TARC Citation Guide EndNote
