Assessment of ROI Selection for Facial Video-Based rPPG
In general, facial image-based remote photoplethysmography (rPPG) methods use color-based and patch-based region-of-interest (ROI) selection methods to estimate the blood volume pulse (BVP) and beats per minute (BPM). Anatomically, the thickness of the skin is not uniform in all areas of the face, s...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/23/7923 |
| _version_ | 1797507195018936320 |
|---|---|
| author | Dae-Yeol Kim Kwangkee Lee Chae-Bong Sohn |
| author_facet | Dae-Yeol Kim Kwangkee Lee Chae-Bong Sohn |
| author_sort | Dae-Yeol Kim |
| collection | DOAJ |
| description | In general, facial image-based remote photoplethysmography (rPPG) methods use color-based and patch-based region-of-interest (ROI) selection methods to estimate the blood volume pulse (BVP) and beats per minute (BPM). Anatomically, the thickness of the skin is not uniform in all areas of the face, so the same diffuse reflection information cannot be obtained in each area. In recent years, various studies have presented experimental results for their ROIs but did not provide a valid rationale for the proposed regions. In this paper, to see the effect of skin thickness on the accuracy of the rPPG algorithm, we conducted an experiment on 39 anatomically divided facial regions. Experiments were performed with seven algorithms (CHROM, GREEN, ICA, PBV, POS, SSR, and LGI) using the UBFC-rPPG and LGI-PPGI datasets considering 29 selected regions and two adjusted regions out of 39 anatomically classified regions. We proposed a BVP similarity evaluation metric to find a region with high accuracy. We conducted additional experiments on the TOP-5 regions and BOT-5 regions and presented the validity of the proposed ROIs. The TOP-5 regions showed relatively high accuracy compared to the previous algorithm’s ROI, suggesting that the anatomical characteristics of the ROI should be considered when developing a facial image-based rPPG algorithm. |
| first_indexed | 2024-03-10T04:45:09Z |
| format | Article |
| id | doaj.art-240207dd9e3646f59a103b25757b30dc |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T04:45:09Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-240207dd9e3646f59a103b25757b30dc2023-11-23T03:01:15ZengMDPI AGSensors1424-82202021-11-012123792310.3390/s21237923Assessment of ROI Selection for Facial Video-Based rPPGDae-Yeol Kim0Kwangkee Lee1Chae-Bong Sohn2Tvstorm, Sunghyun Building, 255 Hyorung-to, Secho-gu, Seoul 13875, KoreaTvstorm, Sunghyun Building, 255 Hyorung-to, Secho-gu, Seoul 13875, KoreaDepartment of Electronics and Communications Engineering, Kwangwoon University, Seoul 01897, KoreaIn general, facial image-based remote photoplethysmography (rPPG) methods use color-based and patch-based region-of-interest (ROI) selection methods to estimate the blood volume pulse (BVP) and beats per minute (BPM). Anatomically, the thickness of the skin is not uniform in all areas of the face, so the same diffuse reflection information cannot be obtained in each area. In recent years, various studies have presented experimental results for their ROIs but did not provide a valid rationale for the proposed regions. In this paper, to see the effect of skin thickness on the accuracy of the rPPG algorithm, we conducted an experiment on 39 anatomically divided facial regions. Experiments were performed with seven algorithms (CHROM, GREEN, ICA, PBV, POS, SSR, and LGI) using the UBFC-rPPG and LGI-PPGI datasets considering 29 selected regions and two adjusted regions out of 39 anatomically classified regions. We proposed a BVP similarity evaluation metric to find a region with high accuracy. We conducted additional experiments on the TOP-5 regions and BOT-5 regions and presented the validity of the proposed ROIs. The TOP-5 regions showed relatively high accuracy compared to the previous algorithm’s ROI, suggesting that the anatomical characteristics of the ROI should be considered when developing a facial image-based rPPG algorithm.https://www.mdpi.com/1424-8220/21/23/7923remote photoplethysmography(rPPG)facial image-based ROI selectionBVP similarity |
| spellingShingle | Dae-Yeol Kim Kwangkee Lee Chae-Bong Sohn Assessment of ROI Selection for Facial Video-Based rPPG Sensors remote photoplethysmography(rPPG) facial image-based ROI selection BVP similarity |
| title | Assessment of ROI Selection for Facial Video-Based rPPG |
| title_full | Assessment of ROI Selection for Facial Video-Based rPPG |
| title_fullStr | Assessment of ROI Selection for Facial Video-Based rPPG |
| title_full_unstemmed | Assessment of ROI Selection for Facial Video-Based rPPG |
| title_short | Assessment of ROI Selection for Facial Video-Based rPPG |
| title_sort | assessment of roi selection for facial video based rppg |
| topic | remote photoplethysmography(rPPG) facial image-based ROI selection BVP similarity |
| url | https://www.mdpi.com/1424-8220/21/23/7923 |
| work_keys_str_mv | AT daeyeolkim assessmentofroiselectionforfacialvideobasedrppg AT kwangkeelee assessmentofroiselectionforfacialvideobasedrppg AT chaebongsohn assessmentofroiselectionforfacialvideobasedrppg |