MR-NIRP Car Dataset
收藏资源简介:
MR-NIRP Car Dataset是由埃因霍温理工大学和Philips Intellectual Property & Standards创建的,用于评估在驾驶环境中使用近红外(NIR)摄像头进行图像光电容积描记(iPPG)的心率监测性能。该数据集包含18名驾驶员在不同驾驶条件下的面部视频,视频帧率为30fps,并使用940 ± 5 nm带通滤波器进行过滤。数据集的创建过程包括使用MediaPipe进行面部特征提取和选择23个感兴趣区域(ROIs),以及通过CMS 50D+指尖脉搏血氧仪收集地面真值脉搏波形。该数据集主要应用于驾驶员监控系统,旨在通过连续监测驾驶员的心率来减少交通事故。
MR-NIRP Car Dataset was developed by Eindhoven University of Technology and Philips Intellectual Property & Standards, for evaluating the heart rate monitoring performance of imaging photoplethysmography (iPPG) using near-infrared (NIR) cameras in driving environments. This dataset includes facial videos of 18 drivers under different driving conditions, with a frame rate of 30 fps, and filtered by a 940 ± 5 nm band-pass filter. The dataset construction process involves facial feature extraction using MediaPipe and selection of 23 regions of interest (ROIs), as well as collection of ground-truth pulse waveforms via a CMS 50D+ fingertip pulse oximeter. This dataset is primarily applied to driver monitoring systems, aiming to reduce traffic accidents by continuously monitoring the driver's heart rate.

- 1Internship Report: Benchmark of Deep Learning-based Imaging PPG in Automotive Domain埃因霍温理工大学 · 2024年



