IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN)
Unmanned Aerial Vehicle (UAV) missions often collect large volumes of imagery data. However, not all images will have useful information, or be of sufficient quality. Manually sorting these images and selecting useful data are both time consuming and prone to interpreter bias. Deep neural network al...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2020-11-01
|
| Series: | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIV-M-2-2020/99/2020/isprs-archives-XLIV-M-2-2020-99-2020.pdf |
| _version_ | 1828811402637213696 |
|---|---|
| author | A. Shashank V. V. Sajithvariyar V. Sowmya K. P. Soman R. Sivanpillai G. K. Brown |
| author_facet | A. Shashank V. V. Sajithvariyar V. Sowmya K. P. Soman R. Sivanpillai G. K. Brown |
| author_sort | A. Shashank |
| collection | DOAJ |
| description | Unmanned Aerial Vehicle (UAV) missions often collect large volumes of imagery data. However, not all images will have useful information, or be of sufficient quality. Manually sorting these images and selecting useful data are both time consuming and prone to interpreter bias. Deep neural network algorithms are capable of processing large image datasets and can be trained to identify specific targets. Generative Adversarial Networks (GANs) consist of two competing networks, <i>Generator</i> and <i>Discriminator</i> that can analyze, capture, and copy the variations within a given dataset. In this study, we selected a variant of GAN called Conditional-GAN that incorporates an additional label parameter, for identifying epiphytes in photos acquired by a UAV in forests within Costa Rica. We trained the network with 70%, 80%, and 90% of 119 photos containing the target epiphyte, <i>Werauhia kupperiana</i> (Bromeliaceae) and validated the algorithm’s performance using a validation data that were not used for training. The accuracy of the output was measured using structural similarity index measure (SSIM) index and histogram correlation (HC) coefficient. Results obtained in this study indicated that the output images generated by C-GAN were similar (average SSIM = 0.89–0.91 and average HC 0.97–0.99) to the analyst annotated images. However, C-GAN had difficulty to identify when the target plant was away from the camera, was not well lit, or covered by other plants. Results obtained in this study demonstrate the potential of C-GAN to reduce the time spent by botanists to identity epiphytes in images acquired by UAVs. |
| first_indexed | 2024-12-12T09:27:17Z |
| format | Article |
| id | doaj.art-1b7eed3fa28849369292d3946b2cc52e |
| institution | Directory Open Access Journal |
| issn | 1682-1750 2194-9034 |
| language | English |
| last_indexed | 2024-12-12T09:27:17Z |
| publishDate | 2020-11-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| spelling | doaj.art-1b7eed3fa28849369292d3946b2cc52e2022-12-22T00:29:00ZengCopernicus PublicationsThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences1682-17502194-90342020-11-01XLIV-M-2-20209910410.5194/isprs-archives-XLIV-M-2-2020-99-2020IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN)A. Shashank0V. V. Sajithvariyar1V. Sowmya2K. P. Soman3R. Sivanpillai4G. K. Brown5Center for Computational Engineering and Networking, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore, TN 641 112, IndiaCenter for Computational Engineering and Networking, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore, TN 641 112, IndiaCenter for Computational Engineering and Networking, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore, TN 641 112, IndiaCenter for Computational Engineering and Networking, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore, TN 641 112, IndiaWyoming GIS Center, University of Wyoming, Laramie, WY 82072, USADepartment of Botany, University of Wyoming, Laramie, WY 82072, USAUnmanned Aerial Vehicle (UAV) missions often collect large volumes of imagery data. However, not all images will have useful information, or be of sufficient quality. Manually sorting these images and selecting useful data are both time consuming and prone to interpreter bias. Deep neural network algorithms are capable of processing large image datasets and can be trained to identify specific targets. Generative Adversarial Networks (GANs) consist of two competing networks, <i>Generator</i> and <i>Discriminator</i> that can analyze, capture, and copy the variations within a given dataset. In this study, we selected a variant of GAN called Conditional-GAN that incorporates an additional label parameter, for identifying epiphytes in photos acquired by a UAV in forests within Costa Rica. We trained the network with 70%, 80%, and 90% of 119 photos containing the target epiphyte, <i>Werauhia kupperiana</i> (Bromeliaceae) and validated the algorithm’s performance using a validation data that were not used for training. The accuracy of the output was measured using structural similarity index measure (SSIM) index and histogram correlation (HC) coefficient. Results obtained in this study indicated that the output images generated by C-GAN were similar (average SSIM = 0.89–0.91 and average HC 0.97–0.99) to the analyst annotated images. However, C-GAN had difficulty to identify when the target plant was away from the camera, was not well lit, or covered by other plants. Results obtained in this study demonstrate the potential of C-GAN to reduce the time spent by botanists to identity epiphytes in images acquired by UAVs.https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIV-M-2-2020/99/2020/isprs-archives-XLIV-M-2-2020-99-2020.pdf |
| spellingShingle | A. Shashank V. V. Sajithvariyar V. Sowmya K. P. Soman R. Sivanpillai G. K. Brown IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| title | IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) |
| title_full | IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) |
| title_fullStr | IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) |
| title_full_unstemmed | IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) |
| title_short | IDENTIFYING EPIPHYTES IN DRONES PHOTOS WITH A CONDITIONAL GENERATIVE ADVERSARIAL NETWORK (C-GAN) |
| title_sort | identifying epiphytes in drones photos with a conditional generative adversarial network c gan |
| url | https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLIV-M-2-2020/99/2020/isprs-archives-XLIV-M-2-2020-99-2020.pdf |
| work_keys_str_mv | AT ashashank identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan AT vvsajithvariyar identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan AT vsowmya identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan AT kpsoman identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan AT rsivanpillai identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan AT gkbrown identifyingepiphytesindronesphotoswithaconditionalgenerativeadversarialnetworkcgan |