Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data
Abstract Many studies of Earth surface processes and landscape evolution rely on having accurate and extensive data sets of surficial geologic units and landforms. Automated extraction of geomorphic features using deep learning provides an objective way to consistently map landforms over large spati...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
American Geophysical Union (AGU)
2023-05-01
|
| Series: | Earth and Space Science |
| Subjects: | |
| Online Access: | https://doi.org/10.1029/2023EA002845 |
| _version_ | 1797820090940391424 |
|---|---|
| author | Aaron E. Maxwell William E. Odom Charles M. Shobe Daniel H. Doctor Michelle S. Bester Tobi Ore |
| author_facet | Aaron E. Maxwell William E. Odom Charles M. Shobe Daniel H. Doctor Michelle S. Bester Tobi Ore |
| author_sort | Aaron E. Maxwell |
| collection | DOAJ |
| description | Abstract Many studies of Earth surface processes and landscape evolution rely on having accurate and extensive data sets of surficial geologic units and landforms. Automated extraction of geomorphic features using deep learning provides an objective way to consistently map landforms over large spatial extents. However, there is no consensus on the optimal input feature space for such analyses. We explore the impact of input feature space for extracting geomorphic features from land surface parameters (LSPs) derived from digital terrain models (DTMs) using convolutional neural network (CNN)‐based semantic segmentation deep learning. We compare four input feature space configurations: (a) a three‐layer composite consisting of a topographic position index (TPI) calculated using a 50 m radius circular window, square root of topographic slope, and TPI calculated using an annulus with a 2 m inner radius and 10 m outer radius, (b) a single illuminating position hillshade, (c) a multidirectional hillshade, and (d) a slopeshade. We test each feature space input using three deep learning algorithms and four use cases: two with natural features and two with anthropogenic features. The three‐layer composite generally provided lower overall losses for the training samples, a higher F1‐score for the withheld validation data, and better performance for generalizing to withheld testing data from a new geographic extent. Results suggest that CNN‐based deep learning for mapping geomorphic features or landforms from LSPs is sensitive to input feature space. Given the large number of LSPs that can be derived from DTM data and the variety of geomorphic mapping tasks that can be undertaken using CNN‐based methods, we argue that additional research focused on feature space considerations is needed and suggest future research directions. We also suggest that the three‐layer composite implemented here can offer better performance in comparison to using hillshades or other common terrain visualization surfaces and is, thus, worth considering for different mapping and feature extraction tasks. |
| first_indexed | 2024-03-13T09:32:15Z |
| format | Article |
| id | doaj.art-4d0feb0f39a141d5a13fd29a5d89436d |
| institution | Directory Open Access Journal |
| issn | 2333-5084 |
| language | English |
| last_indexed | 2024-03-13T09:32:15Z |
| publishDate | 2023-05-01 |
| publisher | American Geophysical Union (AGU) |
| record_format | Article |
| series | Earth and Space Science |
| spelling | doaj.art-4d0feb0f39a141d5a13fd29a5d89436d2023-05-25T20:18:32ZengAmerican Geophysical Union (AGU)Earth and Space Science2333-50842023-05-01105n/an/a10.1029/2023EA002845Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain DataAaron E. Maxwell0William E. Odom1Charles M. Shobe2Daniel H. Doctor3Michelle S. Bester4Tobi Ore5Department of Geology and Geography West Virginia University Morgantown WV USAFlorence Bascom Geoscience Center U.S. Geological Survey Reston VA USADepartment of Geology and Geography West Virginia University Morgantown WV USAFlorence Bascom Geoscience Center U.S. Geological Survey Reston VA USADepartment of Geology and Geography West Virginia University Morgantown WV USADepartment of Geology and Geography West Virginia University Morgantown WV USAAbstract Many studies of Earth surface processes and landscape evolution rely on having accurate and extensive data sets of surficial geologic units and landforms. Automated extraction of geomorphic features using deep learning provides an objective way to consistently map landforms over large spatial extents. However, there is no consensus on the optimal input feature space for such analyses. We explore the impact of input feature space for extracting geomorphic features from land surface parameters (LSPs) derived from digital terrain models (DTMs) using convolutional neural network (CNN)‐based semantic segmentation deep learning. We compare four input feature space configurations: (a) a three‐layer composite consisting of a topographic position index (TPI) calculated using a 50 m radius circular window, square root of topographic slope, and TPI calculated using an annulus with a 2 m inner radius and 10 m outer radius, (b) a single illuminating position hillshade, (c) a multidirectional hillshade, and (d) a slopeshade. We test each feature space input using three deep learning algorithms and four use cases: two with natural features and two with anthropogenic features. The three‐layer composite generally provided lower overall losses for the training samples, a higher F1‐score for the withheld validation data, and better performance for generalizing to withheld testing data from a new geographic extent. Results suggest that CNN‐based deep learning for mapping geomorphic features or landforms from LSPs is sensitive to input feature space. Given the large number of LSPs that can be derived from DTM data and the variety of geomorphic mapping tasks that can be undertaken using CNN‐based methods, we argue that additional research focused on feature space considerations is needed and suggest future research directions. We also suggest that the three‐layer composite implemented here can offer better performance in comparison to using hillshades or other common terrain visualization surfaces and is, thus, worth considering for different mapping and feature extraction tasks.https://doi.org/10.1029/2023EA002845geomorphic mappingconvolutional neural networkssemantic segmentationdigital elevation datadigital terrain datadeep learning |
| spellingShingle | Aaron E. Maxwell William E. Odom Charles M. Shobe Daniel H. Doctor Michelle S. Bester Tobi Ore Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data Earth and Space Science geomorphic mapping convolutional neural networks semantic segmentation digital elevation data digital terrain data deep learning |
| title | Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data |
| title_full | Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data |
| title_fullStr | Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data |
| title_full_unstemmed | Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data |
| title_short | Exploring the Influence of Input Feature Space on CNN‐Based Geomorphic Feature Extraction From Digital Terrain Data |
| title_sort | exploring the influence of input feature space on cnn based geomorphic feature extraction from digital terrain data |
| topic | geomorphic mapping convolutional neural networks semantic segmentation digital elevation data digital terrain data deep learning |
| url | https://doi.org/10.1029/2023EA002845 |
| work_keys_str_mv | AT aaronemaxwell exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata AT williameodom exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata AT charlesmshobe exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata AT danielhdoctor exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata AT michellesbester exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata AT tobiore exploringtheinfluenceofinputfeaturespaceoncnnbasedgeomorphicfeatureextractionfromdigitalterraindata |