Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model

Abstract Snow pole time‐lapse photography, in which a snow pole of a known height is installed in front of a camera and photographed repeatedly over the course of a snow season, allows a large network of sites to be established relatively quickly and affordably. However, current approaches for extra...

Full description

Saved in:
Bibliographic Details
Main Authors: C. M. Breen, W. R. Currier, C. Vuyovich, Z. Miao, L. R. Prugh
Format: Article
Language:English
Published: Wiley 2024-07-01
Series:Water Resources Research
Subjects:
Online Access:https://doi.org/10.1029/2023WR036682
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850115183873097728
author C. M. Breen
W. R. Currier
C. Vuyovich
Z. Miao
L. R. Prugh
author_facet C. M. Breen
W. R. Currier
C. Vuyovich
Z. Miao
L. R. Prugh
author_sort C. M. Breen
collection DOAJ
description Abstract Snow pole time‐lapse photography, in which a snow pole of a known height is installed in front of a camera and photographed repeatedly over the course of a snow season, allows a large network of sites to be established relatively quickly and affordably. However, current approaches for extracting snow depth from snow poles typically relies on time intensive manual photo processing. By integrating computer vision algorithms with snow pole photography, we present a method that uses a keypoint detection model to automatically observe snow height across a network of sites. At 20 snow pole locations from Grand Mesa, CO (n = 9,722 images), our model successfully predicts the top and bottom of the pole with a mean absolute error (MAE) of 1.30 cm. To assess model generalizability, we tested the model on 12 sites in Washington State (n = 1,770 images). When the Colorado trained model was fine‐tuned using a subset of Washington images, the model predicted snow depth with a MAE of 4.0 cm. Best performance was achieved when both data sets were included during training, with a MAE of 2.05 cm for Colorado images and a MAE of 1.14 cm for Washington images. We demonstrate that, especially when trained using a subset of site‐specific data, a keypoint detection model can accelerate snow pole automation. This algorithm brings the hydrology community one step closer to a generalized snow pole detection model, and we call for a future model that integrates across time‐lapse images from additional locations.
format Article
id doaj-art-1587a4bd7c9e42fdbd5b4f2262a611e7
institution OA Journals
issn 0043-1397
1944-7973
language English
publishDate 2024-07-01
publisher Wiley
record_format Article
series Water Resources Research
spelling doaj-art-1587a4bd7c9e42fdbd5b4f2262a611e72025-08-20T02:36:39ZengWileyWater Resources Research0043-13971944-79732024-07-01607n/an/a10.1029/2023WR036682Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning ModelC. M. Breen0W. R. Currier1C. Vuyovich2Z. Miao3L. R. Prugh4Department of Environmental and Forest Sciences University of Washington Seattle WA USAPhysical Sciences Laboratory National Oceanic and Atmospheric Administration Boulder CO USAHydrological Sciences Laboratory NASA Goddard Space Flight Center Greenbelt MD USAAI For Good Research Lab Microsoft Seattle WA USADepartment of Environmental and Forest Sciences University of Washington Seattle WA USAAbstract Snow pole time‐lapse photography, in which a snow pole of a known height is installed in front of a camera and photographed repeatedly over the course of a snow season, allows a large network of sites to be established relatively quickly and affordably. However, current approaches for extracting snow depth from snow poles typically relies on time intensive manual photo processing. By integrating computer vision algorithms with snow pole photography, we present a method that uses a keypoint detection model to automatically observe snow height across a network of sites. At 20 snow pole locations from Grand Mesa, CO (n = 9,722 images), our model successfully predicts the top and bottom of the pole with a mean absolute error (MAE) of 1.30 cm. To assess model generalizability, we tested the model on 12 sites in Washington State (n = 1,770 images). When the Colorado trained model was fine‐tuned using a subset of Washington images, the model predicted snow depth with a MAE of 4.0 cm. Best performance was achieved when both data sets were included during training, with a MAE of 2.05 cm for Colorado images and a MAE of 1.14 cm for Washington images. We demonstrate that, especially when trained using a subset of site‐specific data, a keypoint detection model can accelerate snow pole automation. This algorithm brings the hydrology community one step closer to a generalized snow pole detection model, and we call for a future model that integrates across time‐lapse images from additional locations.https://doi.org/10.1029/2023WR036682snow depthkeypoint modeltime‐lapse camerasautomation
spellingShingle C. M. Breen
W. R. Currier
C. Vuyovich
Z. Miao
L. R. Prugh
Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
Water Resources Research
snow depth
keypoint model
time‐lapse cameras
automation
title Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
title_full Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
title_fullStr Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
title_full_unstemmed Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
title_short Snow Depth Extraction From Time‐Lapse Imagery Using a Keypoint Deep Learning Model
title_sort snow depth extraction from time lapse imagery using a keypoint deep learning model
topic snow depth
keypoint model
time‐lapse cameras
automation
url https://doi.org/10.1029/2023WR036682
work_keys_str_mv AT cmbreen snowdepthextractionfromtimelapseimageryusingakeypointdeeplearningmodel
AT wrcurrier snowdepthextractionfromtimelapseimageryusingakeypointdeeplearningmodel
AT cvuyovich snowdepthextractionfromtimelapseimageryusingakeypointdeeplearningmodel
AT zmiao snowdepthextractionfromtimelapseimageryusingakeypointdeeplearningmodel
AT lrprugh snowdepthextractionfromtimelapseimageryusingakeypointdeeplearningmodel