bcdemattei/plankton_interaction_videos
收藏Hugging Face2026-03-20 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/bcdemattei/plankton_interaction_videos
下载链接
链接失效反馈官方服务:
资源简介:
---
license: apache-2.0
language:
- en
tags:
- plankton
- ecology
- biology
pretty_name: Predator-Prey Encounter Detection Dataset
size_categories:
- 10K<n<100K
---
Video frame image dataset of bdelloid rotifers (*Philodina* sp.) and cryptomonad algae (*Cryptomonas erosa*). The videos were captured at 25 frames per second using Zeiss ZEN lite and a Zeiss SteREO Discovery v20 dissection microscope with an Axiocam 705 Color under dark-field conditions. A PlanApo S 0.63x objective was used with the zoom set to 120x, which created a field of view with an area of 0.0241mm^2. The frames were extracted from each video and split into Training (80%), Validation (10%), and Testing (10%) datasets to be trained using Ultralytics YOLOv5. In total, 25 videos were taken with a total of 45,158 frames, with the frame count of a given video ranging from 550 to 4000 frames. Text files contain labels in the YOLOv5 format. The labels were originally created using Label Studio.
本数据集相关元信息如下:许可证采用Apache-2.0协议,语言为英语,标注标签涵盖浮游生物、生态学、生物学领域,数据集展示名称为「捕食者-猎物相遇检测数据集」,样本量规模处于10000至100000区间。
本数据集为蛭态轮虫(*Philodina* 属未定种)与啮蚀隐藻(*Cryptomonas erosa*)的视频帧图像数据集。所有视频以25帧每秒的帧率采集,采集流程使用蔡司ZEN lite软件,搭配蔡司SteREO Discovery v20体视显微镜及Axiocam 705 Color彩色相机,拍摄环境为暗场照明条件。实验采用PlanApo S 0.63×物镜,将变焦倍率设置为120×,此时视场面积为0.0241mm²。随后从每段视频中提取帧图像,并按照80%训练集、10%验证集、10%测试集的比例拆分,用于Ultralytics YOLOv5模型的训练。本次共采集25段视频,总帧数达45158帧,单段视频的帧数范围为550至4000帧。配套文本文件采用YOLOv5格式存储标注信息,标注工作最初通过Label Studio工具完成。
提供机构:
bcdemattei



