Papers › Deep learning with RGB and thermal images onboard a drone for monitoring operations

Deep learning with RGB and thermal images onboard a drone for monitoring operations

31 May 2022Journal of Field Robotics 2022 5archive 2025-07-28

Simon Speth, Artur Gonçalves, Bastien Rigault, Satoshi Suzuki, Mondher Bouazizi, Yutaka Matsuo, Helmut Prendinger

This article describes the artificial intelligence (AI) component of a drone for monitoring and patrolling tasks associated with disaster relief missions in specific restricted disaster scenarios, as specified by the Advanced Robotics Foundation in Japan. The AI component uses deep learning models for environment recognition and object detection. For environment recognition, we use semantic segmentation, or pixel-wise labeling, based on RGB images. Object detection is key for detecting and locating people in need. Since people are relatively small objects from the drone perspective, we use both RGB and thermal images. To train our models, we created a novel multispectral and publicly available data set of people. We used a geo-location method to locate people on the ground. The semantic segmentation models were extensively tested using different feature extractors. We created two dedicated data sets, which we have made publicly available. Compared with the baseline model, the best-performing model could increase the mean intersection over union (IoU) by 1.3%. Furthermore, we compared two types of person detection models. The first one is an ensemble model that combines RGB and thermal information via “late fusion”; the second one is a 4-channel model that combines these two types of information in an “early fusion” manner. The results suggest that the 4-channel model had a 40.6% increase of average precision for stricter IoU values (0.75) compared with the ensemble model and a 5.8% increase in the average precision compared with the thermal model. All models were deployed and tested on the NVIDIA AGX Xavier platform. To the best of our knowledge, this study was the first to use both RGB and thermal data from the perspective of a drone for monitoring tasks.

PaperPDF

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Human DetectionMultispectral Object DetectionObject DetectionSemantic Segmentationobject-detection

Datasets

Introduced by this paper, per the archive.

NII-CU MAPDOkutama Drone and Swiss Drone Dataset

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Multispectral Object Detection NII-CU MAPD YOLOv3-4‐channel AP@0.5 97.9 #1 of 2 Archive leaderboard report
Multispectral Object Detection NII-CU MAPD YOLOv3-4‐channel AP@0.75 76.9 #1 of 2 Archive leaderboard report
Multispectral Object Detection NII-CU MAPD YOLOv3-4‐channel mAP@0.5:0.95 64.4 #1 of 2 Archive leaderboard report
Multispectral Object Detection NII-CU MAPD YOLOv3-Ensemble AP@0.5 97.3 #2 of 2 Archive leaderboard report
Multispectral Object Detection NII-CU MAPD YOLOv3-Ensemble AP@0.75 54.7 #2 of 2 Archive leaderboard report
Multispectral Object Detection NII-CU MAPD YOLOv3-Ensemble mAP@0.5:0.95 53.4 #2 of 2 Archive leaderboard report
Object Detection NII-CU MAPD YOLOv3 AP@0.5 92.4 #1 of 1 Archive leaderboard report
Object Detection NII-CU MAPD YOLOv3 AP@0.75 44.5 #1 of 1 Archive leaderboard report
Object Detection NII-CU MAPD YOLOv3 mAP@0.5:0.95 48.3 #1 of 1 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐ResNet‐101 Acc 90.78 #1 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐ResNet‐101 mIoU 65.88 #1 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐Xception‐65 Acc 90.72 #2 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐Xception‐65 mIoU 64.34 #2 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐ResNet‐50 Acc 78.65 #3 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐ResNet‐50 mIoU 43.65 #3 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐Xception‐71 Acc 74.31 #4 of 4 Archive leaderboard report
Semantic Segmentation Okutama Drone and Swiss Drone Dataset DeepLabv3+‐Xception‐71 mIoU 37.81 #4 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections