Influence of Depth Camera Noise Models on Respiration Estimation

TL;DR

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 are shown.

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.

Figures (9)

• Figure 1: Overview of rendering and post-processing pipeline
• Figure 2: Body part mostly affected by thoracic respiration.
• Figure 3: Visualization of applied depth camera noise types. Noise is exaggerated for visualization purposes. a) Gaussian noise, b) axial noise, c) radial noise, d) motion noise (difference image), e) edge permutation noise, f) edge Gaussian noise.
• Figure 4: Select a RoI of the chest to extract signal.
• Figure 5: Fantasia reference signal and extracted noisy signal ($\sigma=0.067, \text{scale}=0.2, \text{SNR}=6.7$)