Non contact pulmonary functional testing through an improved photometric stereo approach

Ahmad, Jahanzeb; Sun, jiuai; Smith, Lyndon; Smith, Melvyn

Abstract

A non-contact computer vision based system is developed for Pulmonary Functional Testing. The unique and novel features of the system are that it views the patients from both front and back and creates a 3D structure of the whole torso. By observing the 3D structure of the torso over time, the amount of air inhaled and exhaled is estimated.
The Photometric Stereo method is used to recover local 3D surface orientation. This technique is good for recovering the local 3D surface orientation but provides no absolute
depth information about the observed scene. To calculate absolute depth information from Photometric Stereo a new Lambertian Diffused Maxima Region (DMR) based algorithm
is employed here. A high (0.96) correlation between breathing volume calculated from Photometric Stereo and Spirometer data is observed.

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: Medical Image Understanding and Analysis 2014
• Start Date: Jul 9, 2014
• End Date: Jul 14, 2014
• Publication Date: Jul 9, 2014
• Publicly Available Date: Jun 6, 2019
• Peer Reviewed: Peer Reviewed
• Keywords: non contact, pulmonary functional testing, improved photometric stereo
• Public URL: https://uwe-repository.worktribe.com/output/814710
• Publisher URL: http://www.city.ac.uk/medical-image-understanding-and-analysis-2014/programme
• Related Public URLs: http://www.city.ac.uk/medical-image-understanding-and-analysis-2014
• Additional Information: Title of Conference or Conference Proceedings : Medical Image Understanding and Analysis 2014