In ecological research, accurately collecting spatiotemporal position data is
a fundamental task for understanding the behavior and ecology of insects and
other organisms. In recent years, advancements in computer vision techniques
have reached a stage of maturity where they can support, and in some cases,
replace manual observation. In this study, a simple and inexpensive method for
monitoring insects in three dimensions (3D) was developed so that their
behavior could be observed automatically in experimental environments. The main
achievements of this study have been to create a 3D monitoring algorithm using
inexpensive cameras and other equipment to design an adjusting algorithm for
depth error, and to validate how our plotting algorithm is quantitatively
precise, all of which had not been realized in conventional studies. By
offering detailed 3D visualizations of insects, the plotting algorithm aids
researchers in more effectively comprehending how insects interact within their
environments.

本研究开发了一种简单且廉价的三维昆虫监测算法，利用廉价相机和其他设备设计了一个深度误差调整算法，通过提供昆虫的详细三维可视化，该绘图算法有助于研究人员更有效地理解昆虫在其环境中的互动。

基于 YOLOv5 的昆虫三维绘图算法

3D-Plotting Algorithm for Insects using YOLOv5

Deep learning methods for computer vision tasks show promise for automating
the data analysis of camera trap images. Ecological camera traps are a common
approach for monitoring an ecosystem's animal population, as they provide
continual insight into an environment without being intrusive. However, the
analysis of camera trap images is expensive, labour intensive, and time
consuming. Recent advances in the field of deep learning for object detection
show promise towards automating the analysis of camera trap images. Here, we
demonstrate their capabilities by training and comparing two deep learning
object detection classifiers, Faster R-CNN and YOLO v2.0, to identify,
quantify, and localize animal species within camera trap images using the
Reconyx Camera Trap and the self-labeled Gold Standard Snapshot Serengeti data
sets. When trained on large labeled datasets, object recognition methods have
shown success. We demonstrate their use, in the context of realistically sized
ecological data sets, by testing if object detection methods are applicable for
ecological research scenarios when utilizing transfer learning. Faster R-CNN
outperformed YOLO v2.0 with average accuracies of 93.0\% and 76.7\% on the two
data sets, respectively. Our findings show promising steps towards the
automation of the labourious task of labeling camera trap images, which can be
used to improve our understanding of the population dynamics of ecosystems
across the planet.

本文论述了如何通过使用深度学习方法对相机陷阱图像进行自动化数据分析，演示了以 Faster R-CNN 和 YOLO v2.0 作为分类器识别、量化和定位相机陷阱图像中的物种，进而应用于生态研究。结果显示 Faster R-CNN 的分类准确率高达 93.0％，取得了很好的效果。