Non-invasive detection algorithm of thermal comfort based on computer vision

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Master's thesis

Date

2018-08-20

Department

Major/Subject

ICT Innovation - EIT Digital Master School

Mcode

SCI3020

Degree programme

Master's Programme in ICT Innovation

Language

en

Pages

26 + 0

Series

Abstract

The waste of building energy consumption is a major challenge in the world. Real-time detection of human thermal comfort is an effective way to deal with this issue. However, due to the difference of personal thermal comfort and changes caused by climatic variations, there is still a long way to reach this target. From another perspective, the current HVAC (heating, ventilating and air-conditioning) systems are reluctant to provide flexible interaction channels to adjust atmosphere which fails to follow continuously increasing requirements from users. All of them indicate the necessity to develop more intelligent detection method for human thermal comfort. In this paper, a non-invasion detection method toward thermal comfort is proposed from two perspectives: macro human postures and skin textures. In posture part, OpenPose is used for detecting the key points’ position coordinates of human body in images, which would be functionalized from the term of thermal comfort. In skin textures, deep neural network is used to regress the images of skin to its temperature. Based on Fanger’s theory of thermal comfort, the results of both parts are satisfying: subjects’ postures can be captured and interpreted into different thermal comfort level: hot, cold and comfort. And the absolute error of prediction from neurons network is less than 0.125 degrees centigrade which is the equipment error of thermometer used in data acquisition. With solutions of this paper, it is promising to non-invasively detect the thermal comfort level of users from postures and skin textures. And the conclusion and future work are discussed in final chapter.

Description

Supervisor

David, McGookin

Thesis advisor

Xiaogang, Cheng

Keywords

non-invasive, deep learning, OpenPose, computer vision, human thermal comfort

Other note

Citation