# An Autonomous Drone for Search and Rescue in Forests using Airborne Optical Sectioning arXiv: https://arxiv.org/abs/2105.04328 ยท HTML: https://mlsanity.com/paper/2105.04328/ ## Key facts - Cited by 32 papers in the map (data through 14 Oct 2026). - Builds on 10 of its 87 references that are in the map. - First posted on arXiv 10 May 2021. ## Builds on (most cited first, 10 of 10) - [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014) - [You Only Look Once: Unified, Real-Time Object Detection](https://mlsanity.com/paper/1506.02640/) (2015) - [YOLOv4: Optimal Speed and Accuracy of Object Detection](https://mlsanity.com/paper/2004.10934/) (2020) - [Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition](https://mlsanity.com/paper/1406.4729/) (2014) - [Unmanned Aerial Vehicles: A Survey on Civil Applications and Key Research Challenges](https://mlsanity.com/paper/1805.00881/) (2018) - [Search and Rescue with Airborne Optical Sectioning](https://mlsanity.com/paper/2009.08835/) (2020) - [A Statistical View on Synthetic Aperture Imaging for Occlusion Removal](https://mlsanity.com/paper/1906.06600/) (2019) - [DC-SPP-YOLO: Dense Connection and Spatial Pyramid Pooling Based YOLO for Object Detection](https://mlsanity.com/paper/1903.08589/) (2019) - [Fast Automatic Visibility Optimization for Thermal Synthetic Aperture Visualization](https://mlsanity.com/paper/2005.04065/) (2020) - [Pose Error Reduction for Focus Enhancement in Thermal Synthetic Aperture Visualization](https://mlsanity.com/paper/2012.08606/) (2020) ## Built on by (most cited first, 30 of 32) - [Combined Person Classification with Airborne Optical Sectioning](https://mlsanity.com/paper/2106.10077/) (2021) - [Through-Foliage Tracking with Airborne Optical Sectioning](https://mlsanity.com/paper/2111.06959/) (2021) - [AI-Powered Autonomous Weapons Risk Geopolitical Instability and Threaten AI Research](https://mlsanity.com/paper/2405.01859/) (2024) - [Inverse Airborne Optical Sectioning](https://mlsanity.com/paper/2207.13344/) (2022) - [Evaluation of Color Anomaly Detection in Multispectral Images For Synthetic Aperture Sensing](https://mlsanity.com/paper/2211.04293/) (2022) - [Defining and Evaluating Physical Safety for Large Language Models](https://mlsanity.com/paper/2411.02317/) (2024) - [Visual Attention Prediction Improves Performance of Autonomous Drone Racing Agents](https://mlsanity.com/paper/2201.02569/) (2022) - [AirScape: An Aerial Generative World Model with Motion Controllability](https://mlsanity.com/paper/2507.08885/) (2025) - [On the Role of Field of View for Occlusion Removal with Airborne Optical Sectioning](https://mlsanity.com/paper/2204.13371/) (2022) - [Open-World Drone Active Tracking with Goal-Centered Rewards](https://mlsanity.com/paper/2412.00744/) (2024) - [Fusion of Single and Integral Multispectral Aerial Images](https://mlsanity.com/paper/2311.17515/) (2023) - [AirHunt: Bridging VLM Semantics and Continuous Planning for Efficient Aerial Object Navigation](https://mlsanity.com/paper/2601.12742/) (2026) - [Stereoscopic Depth Perception Through Foliage](https://mlsanity.com/paper/2310.16120/) (2023) - [An Autonomous Drone Swarm for Detecting and Tracking Anomalies among Dense Vegetation](https://mlsanity.com/paper/2407.10754/) (2024) - [DeepForest: Sensing Into Self-Occluding Volumes of Vegetation With Aerial Imaging](https://mlsanity.com/paper/2502.02171/) (2025) - [Event Spectroscopy: Event-based Multispectral and Depth Sensing using Structured Light](https://mlsanity.com/paper/2509.06741/) (2025) - [Instance-level Visual Active Tracking with Occlusion-Aware Planning](https://mlsanity.com/paper/2604.21453/) (2026) - [A Search and Detection Autonomous Drone System: from Design to Implementation](https://mlsanity.com/paper/2211.15866/) (2022) - [Through-Foliage Surface-Temperature Reconstruction for Early Wildfire Detection](https://mlsanity.com/paper/2511.12572/) (2025) - [Synthetic Aperture Sensing for Occlusion Removal with Drone Swarms](https://mlsanity.com/paper/2212.14692/) (2022) - [Experimental validation of UAV search and detection system in real wilderness environment](https://mlsanity.com/paper/2502.17372/) (2025) - [How Robot Dogs See the Unseeable: Improving Visual Interpretability via Peering for Exploratory Robots](https://mlsanity.com/paper/2511.16262/) (2025) - [Synthetic Aperture Anomaly Imaging](https://mlsanity.com/paper/2304.13590/) (2023) - [Reciprocal Visibility](https://mlsanity.com/paper/2402.06991/) (2024) - [Design and Flight Demonstration of a Quadrotor for Urban Mapping and Target Tracking Research](https://mlsanity.com/paper/2402.13195/) (2024) - [MEXGEN: An Effective and Efficient Information Gain Approximation for Information Gathering Path Planning](https://mlsanity.com/paper/2405.02605/) (2024) - [A highly maneuverable flying squirrel drone with agility-improving foldable wings](https://mlsanity.com/paper/2504.09609/) (2025) - [An aerial color image anomaly dataset for search missions in complex forested terrain](https://mlsanity.com/paper/2507.15492/) (2025) - [Under-Canopy Terrain Reconstruction in Dense Forests Using RGB Imaging and Neural 3D Reconstruction](https://mlsanity.com/paper/2601.22861/) (2026) - [Training-Free Open-Vocabulary Visual Grounding for Remote Sensing Images and Videos](https://mlsanity.com/paper/2606.16124/) (2026) ## Abstract (arXiv) Drones will play an essential role in human-machine teaming in future search and rescue (SAR) missions. We present a first prototype that finds people fully autonomously in densely occluded forests. In the course of 17 field experiments conducted over various forest types and under different flying conditions, our drone found 38 out of 42 hidden persons; average precision was 86% for predefined flight paths, while adaptive path planning (where potential findings are double-checked) increased confidence by 15%. Image processing, classification, and dynamic flight-path adaptation are computed onboard in real-time and while flying. Our finding that deep-learning-based person classification is unaffected by sparse and error-prone sampling within one-dimensional synthetic apertures allows flights to be shortened and reduces recording requirements to one-tenth of the number of images needed for sampling using two-dimensional synthetic apertures. The goal of our adaptive path planning is to find people as reliably and quickly as possible, which is essential in time-critical applications, such as SAR. Our drone enables SAR operations in remote areas without stable network coverage, as it transmits to the rescue team only classification results that indicate detections and can thus operate with intermittent minimal-bandwidth connections (e.g., by satellite). Once received, these results can be visually enhanced for interpretation on remote mobile devices.