A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data
Forest damage due to storms causes economic loss and requires a fast response to prevent further damage such as bark beetle infestations. By using Convolutional Neural Networks (CNNs) in conjunction with a GIS, we aim at completely streamlining the detection and mapping process for forest agencies....
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/12/13/2121 |
| _version_ | 1797563476850245632 |
|---|---|
| author | Wolfgang Deigele Melanie Brandmeier Christoph Straub |
| author_facet | Wolfgang Deigele Melanie Brandmeier Christoph Straub |
| author_sort | Wolfgang Deigele |
| collection | DOAJ |
| description | Forest damage due to storms causes economic loss and requires a fast response to prevent further damage such as bark beetle infestations. By using Convolutional Neural Networks (CNNs) in conjunction with a GIS, we aim at completely streamlining the detection and mapping process for forest agencies. We developed and tested different CNNs for rapid windthrow detection based on PlanetScope satellite data and high-resolution aerial image data. Depending on the meteorological situation after the storm, PlanetScope data might be rapidly available due to its high temporal resolution, while the acquisition of high-resolution airborne data often takes weeks to a month and is, therefore, used in a second step for more detailed mapping. The study area is located in Bavaria, Germany (ca. 165 km<sup>2</sup>), and labels for damaged areas were provided by the Bavarian State Institute of Forestry (LWF). Modifications of a U-Net architecture were compared to other approaches using transfer learning (e.g., VGG19) to find the most efficient architecture for the task on both datasets while keeping the computational time low. A custom implementation of U-Net proved to be more accurate than transfer learning, especially on medium (3 m) resolution PlanetScope imagery (intersection over union score (IoU) 0.55) where transfer learning completely failed. Results for transfer learning based on VGG19 on high-resolution aerial image data are comparable to results from the custom U-Net architecture (IoU 0.76 vs. 0.73). When using both architectures on a dataset from a different area (located in Hesse, Germany), however, we find that the custom implementations have problems generalizing on aerial image data while VGG19 still detects most damage in these images. For PlanetScope data, VGG19 again fails while U-Net achieves reasonable mappings. Results highlight the potential of Deep Learning algorithms to detect damaged areas with an IoU of 0.73 on airborne data and 0.55 on Planet Dove data. The proposed workflow with complete integration into ArcGIS is well-suited for rapid first assessments after a storm event that allows for better planning of the flight campaign followed by detailed mapping in a second stage. |
| first_indexed | 2024-03-10T18:43:12Z |
| format | Article |
| id | doaj.art-09865fba375943e48d71b4e0b42587b4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T18:43:12Z |
| publishDate | 2020-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-09865fba375943e48d71b4e0b42587b42023-11-20T05:38:39ZengMDPI AGRemote Sensing2072-42922020-07-011213212110.3390/rs12132121A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image DataWolfgang Deigele0Melanie Brandmeier1Christoph Straub2Esri Germany and Switzerland, Ringstr. 7, 85402 Kranzberg, GermanyEsri Germany and Switzerland, Ringstr. 7, 85402 Kranzberg, GermanyDepartment of Information Technology, Bavarian State Institute of Forestry, Hans Carl-von-Carlowitz-Platz 1, 85354 Freising, GermanyForest damage due to storms causes economic loss and requires a fast response to prevent further damage such as bark beetle infestations. By using Convolutional Neural Networks (CNNs) in conjunction with a GIS, we aim at completely streamlining the detection and mapping process for forest agencies. We developed and tested different CNNs for rapid windthrow detection based on PlanetScope satellite data and high-resolution aerial image data. Depending on the meteorological situation after the storm, PlanetScope data might be rapidly available due to its high temporal resolution, while the acquisition of high-resolution airborne data often takes weeks to a month and is, therefore, used in a second step for more detailed mapping. The study area is located in Bavaria, Germany (ca. 165 km<sup>2</sup>), and labels for damaged areas were provided by the Bavarian State Institute of Forestry (LWF). Modifications of a U-Net architecture were compared to other approaches using transfer learning (e.g., VGG19) to find the most efficient architecture for the task on both datasets while keeping the computational time low. A custom implementation of U-Net proved to be more accurate than transfer learning, especially on medium (3 m) resolution PlanetScope imagery (intersection over union score (IoU) 0.55) where transfer learning completely failed. Results for transfer learning based on VGG19 on high-resolution aerial image data are comparable to results from the custom U-Net architecture (IoU 0.76 vs. 0.73). When using both architectures on a dataset from a different area (located in Hesse, Germany), however, we find that the custom implementations have problems generalizing on aerial image data while VGG19 still detects most damage in these images. For PlanetScope data, VGG19 again fails while U-Net achieves reasonable mappings. Results highlight the potential of Deep Learning algorithms to detect damaged areas with an IoU of 0.73 on airborne data and 0.55 on Planet Dove data. The proposed workflow with complete integration into ArcGIS is well-suited for rapid first assessments after a storm event that allows for better planning of the flight campaign followed by detailed mapping in a second stage.https://www.mdpi.com/2072-4292/12/13/2121forest damage assessmentwindthrowconvolutional neural networksGISremote sensing |
| spellingShingle | Wolfgang Deigele Melanie Brandmeier Christoph Straub A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data Remote Sensing forest damage assessment windthrow convolutional neural networks GIS remote sensing |
| title | A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data |
| title_full | A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data |
| title_fullStr | A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data |
| title_full_unstemmed | A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data |
| title_short | A Hierarchical Deep-Learning Approach for Rapid Windthrow Detection on PlanetScope and High-Resolution Aerial Image Data |
| title_sort | hierarchical deep learning approach for rapid windthrow detection on planetscope and high resolution aerial image data |
| topic | forest damage assessment windthrow convolutional neural networks GIS remote sensing |
| url | https://www.mdpi.com/2072-4292/12/13/2121 |
| work_keys_str_mv | AT wolfgangdeigele ahierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata AT melaniebrandmeier ahierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata AT christophstraub ahierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata AT wolfgangdeigele hierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata AT melaniebrandmeier hierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata AT christophstraub hierarchicaldeeplearningapproachforrapidwindthrowdetectiononplanetscopeandhighresolutionaerialimagedata |