Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery
Road building has long been under-mapped globally, arguably more than any other human activity threatening environmental integrity. Millions of kilometers of unmapped roads have challenged environmental governance and conservation in remote frontiers. Prior attempts to map roads at large scales have...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/16/5/839 |
| _version_ | 1797263982478753792 |
|---|---|
| author | Sean Sloan Raiyan R. Talkhani Tao Huang Jayden Engert William F. Laurance |
| author_facet | Sean Sloan Raiyan R. Talkhani Tao Huang Jayden Engert William F. Laurance |
| author_sort | Sean Sloan |
| collection | DOAJ |
| description | Road building has long been under-mapped globally, arguably more than any other human activity threatening environmental integrity. Millions of kilometers of unmapped roads have challenged environmental governance and conservation in remote frontiers. Prior attempts to map roads at large scales have proven inefficient, incomplete, and unamenable to continuous road monitoring. Recent developments in automated road detection using artificial intelligence have been promising but have neglected the relatively irregular, sparse, rustic roadways characteristic of remote semi-natural areas. In response, we tested the accuracy of automated approaches to large-scale road mapping across remote rural and semi-forested areas of equatorial Asia-Pacific. Three machine learning models based on convolutional neural networks (UNet and two ResNet variants) were trained on road data derived from visual interpretations of freely available high-resolution satellite imagery. The models mapped roads with appreciable accuracies, with F1 scores of 72–81% and intersection over union scores of 43–58%. These results, as well as the purposeful simplicity and availability of our input data, support the possibility of concerted program of exhaustive, automated road mapping and monitoring across large, remote, tropical areas threatened by human encroachment. |
| first_indexed | 2024-04-25T00:21:39Z |
| format | Article |
| id | doaj.art-ee060a82b61f408cac1f81a6627c1c38 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-04-25T00:21:39Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-ee060a82b61f408cac1f81a6627c1c382024-03-12T16:54:11ZengMDPI AGRemote Sensing2072-42922024-02-0116583910.3390/rs16050839Mapping Remote Roads Using Artificial Intelligence and Satellite ImagerySean Sloan0Raiyan R. Talkhani1Tao Huang2Jayden Engert3William F. Laurance4Centre for Tropical Environmental and Sustainability Science, College of Science and Engineering, James Cook University, Cairns, Queensland 4878, AustraliaCollege of Science and Engineering, James Cook University, Cairns, Queensland 4878, AustraliaCollege of Science and Engineering, James Cook University, Cairns, Queensland 4878, AustraliaCentre for Tropical Environmental and Sustainability Science, College of Science and Engineering, James Cook University, Cairns, Queensland 4878, AustraliaCentre for Tropical Environmental and Sustainability Science, College of Science and Engineering, James Cook University, Cairns, Queensland 4878, AustraliaRoad building has long been under-mapped globally, arguably more than any other human activity threatening environmental integrity. Millions of kilometers of unmapped roads have challenged environmental governance and conservation in remote frontiers. Prior attempts to map roads at large scales have proven inefficient, incomplete, and unamenable to continuous road monitoring. Recent developments in automated road detection using artificial intelligence have been promising but have neglected the relatively irregular, sparse, rustic roadways characteristic of remote semi-natural areas. In response, we tested the accuracy of automated approaches to large-scale road mapping across remote rural and semi-forested areas of equatorial Asia-Pacific. Three machine learning models based on convolutional neural networks (UNet and two ResNet variants) were trained on road data derived from visual interpretations of freely available high-resolution satellite imagery. The models mapped roads with appreciable accuracies, with F1 scores of 72–81% and intersection over union scores of 43–58%. These results, as well as the purposeful simplicity and availability of our input data, support the possibility of concerted program of exhaustive, automated road mapping and monitoring across large, remote, tropical areas threatened by human encroachment.https://www.mdpi.com/2072-4292/16/5/839convolutional neural networksroadsremote sensingroad maptropical forestsartificial intelligence |
| spellingShingle | Sean Sloan Raiyan R. Talkhani Tao Huang Jayden Engert William F. Laurance Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery Remote Sensing convolutional neural networks roads remote sensing road map tropical forests artificial intelligence |
| title | Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery |
| title_full | Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery |
| title_fullStr | Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery |
| title_full_unstemmed | Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery |
| title_short | Mapping Remote Roads Using Artificial Intelligence and Satellite Imagery |
| title_sort | mapping remote roads using artificial intelligence and satellite imagery |
| topic | convolutional neural networks roads remote sensing road map tropical forests artificial intelligence |
| url | https://www.mdpi.com/2072-4292/16/5/839 |
| work_keys_str_mv | AT seansloan mappingremoteroadsusingartificialintelligenceandsatelliteimagery AT raiyanrtalkhani mappingremoteroadsusingartificialintelligenceandsatelliteimagery AT taohuang mappingremoteroadsusingartificialintelligenceandsatelliteimagery AT jaydenengert mappingremoteroadsusingartificialintelligenceandsatelliteimagery AT williamflaurance mappingremoteroadsusingartificialintelligenceandsatelliteimagery |