UAS-Based Real-Time Detection of Red-Cockaded Woodpecker Cavities in Heterogeneous Landscapes Using YOLO Object Detection Algorithms
نویسندگان
چکیده
In recent years, deep learning-based approaches have proliferated across a variety of ecological studies. Inspired by learning’s emerging prominence as the preferred tool for analyzing wildlife image datasets, this study employed You Only Look Once (YOLO), single-shot, real-time object detection algorithm, to effectively detect cavity trees Red-cockaded Woodpeckers or RCW (Dryobates borealis). spring 2022, using an unmanned aircraft system (UAS), we conducted presence surveys within 1264-hectare area in Sam Houston National Forest (SHNF). Additionally, known occurrences outside surveyed were aerially photographed, manually annotated, and used training dataset. Both YOLOv4-tiny YOLOv5n architectures selected target models later inferencing separate aerial photos from area. A traditional survey pedestrian methods was also concurrently baseline compare our new methods. Our best-performing model generated mAP (mean Average Precision) 95% F1 score 85% while maintaining inference speed 2.5 frames per second (fps). five unique detected UAS approach, compared with one Model development techniques, such preprocessing images tiling Sliced Aided Hyper Inferencing (SAHI), proved be critical components improved performance. results demonstrated two YOLO SAHI strategies able successfully cavities heavily forested, heterogenous environments semi-automated review. Furthermore, case represents progress towards eventual where managers are targeting small objects. These implications more achievable conservation goals, less costly operations, safer work environment personnel, potentially accurate that difficult
منابع مشابه
Complex-YOLO: Real-time 3D Object Detection on Point Clouds
Lidar based 3D object detection is inevitable for autonomous driving, because it directly links to environmental understanding and therefore builds the base for prediction and motion planning. The capacity of inferencing highly sparse 3D data in real-time is an ill-posed problem for lots of other application areas besides automated vehicles, e.g. augmented reality, personal robotics or industri...
متن کاملRed-cockaded Woodpecker Centers On
T he southern pine beetle, Dendroctonus fiontalis (Coleoptera: Scolytidae), and the red-cockaded woodpecker, Picoides borealis, are indigenous species that occur together in pine (Pinus sp.) forests of the southern United States. For different reasons, these organisms have been the focus of considerable research. Interest in the southern pine beetle centers on the role this insect plays as a pe...
متن کاملAn Experimental Test of Interspecific Competition for Red-cockaded Woodpecker Cavities
Ahntrcrct: To test whether the presence of nest boxes near red-cockaded woodpecker (R(Z\V. Picoirlcs horwh) cavity trees red~~ced ca~it\ USC by other species and irnpro\-ed 1ICW reproductive success on the Francis Marion National Forest in coastal South Carohna, WC ~~XYY~ 3 nest boxes in each of 62 experimental clusters and designated 61 clusters as controls. Our observations of nest box and ca...
متن کاملReal-Time Algorithms of Object Detection Using Classifiers
Object detection, or more generally pattern detection and recognition, can be based on many different principles. The objects can be described through their structure, shape, color, texture, etc. [Blaschko & Lampert (2009); Chen et al. (2004); Fidler & Leonardis (2007); Leibe et al. (2008); Lowe (1999); Serre et al. (2005); Viola & Jones (2001)]; therefore, a variety of object detection mechani...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Remote Sensing
سال: 2023
ISSN: ['2315-4632', '2315-4675']
DOI: https://doi.org/10.3390/rs15040883