Influence of Depth Camera Noise Models on Respiration Estimation

TL;DR

This paper investigates how different depth camera noise models affect the accuracy of respiratory rate estimation, emphasizing the importance of realistic noise simulation for training and testing in multi-camera setups.

Contribution

It introduces a 3D-rendering simulation pipeline that models depth camera noise to generate realistic respiratory signals for algorithm development.

Findings

Gaussian noise models are generally sufficient for high-resolution data

Low-resolution images reveal significant differences between noise models

Realistic noise modeling improves the robustness of respiratory signal extraction

Abstract

Depth cameras are an interesting modality for capturing vital signs such as respiratory rate. Plenty approaches exist to extract vital signs in a controlled setting, but in order to apply them more flexibly for example in multi-camera settings, a simulated environment is needed to generate enough data for training and testing of new algorithms. We show first results of a 3D-rendering simulation pipeline that focuses on different noise models in order to generate realistic, depth-camera based respiratory signals using both synthetic and real respiratory signals as a baseline. While most noise can be accurately modelled as Gaussian in this context, we can show that as soon as the available image resolution is too low, the differences between different noise models surface.