遇见数据集

Data from: Multi-image flock size estimation with CountEm: A case study with half a million Common Eiders and Greater Snow Geese.

收藏
Zenodo2022-01-08 更新2026-05-28 收录
数据链接:
官方服务:

资源简介:

The present data set is related to the manuscript "Multi-image flock size estimation with CountEm: A case study with half a million Common Eiders and Greater Snow Geese" submitted to the Ecosphere journal. The files COEI_data.csv and GSGO_data.csv contain data and results of the 179 COEI and 99 GSGO ECA Flocks images respectively. The files have one row per image. Both files have 9 columns corresponding to the following variables: “Filenumber” is used to identify the number of the corresponding image, namely “001” to “099” for GSGO, and “001” to “179” for COEI. “N”: Total number of annotated birds in the image. “Nest”: Bird number estimation (bN 439 ) obtained with a single real mode CountEm run. “CE_sim”: Empirical coefficient of error (relative standard error, CEe(bN 440 )) obtained from 2000 simulated measurements in simulation mode. “f”: Sampling fraction used in the real mode CountEm run. “n0”: Initial number of quadrats used in the real mode CountEm run. “Q”: Total number of birds (i.e. sample size) counted in the real mode CountEm estimation. “n”: Number of non-empty quadrats counted in the real mode CountEm estimation. “UserTime”: Counting time (in minutes) in the real mode CountEm estimation. “ManualTime”: Counting time (in minutes) of the manual annotation process with ImageJ.

本数据集关联于一篇提交至《生态圈》(Ecosphere)期刊的稿件:《基于CountEm的多图像种群规模估算:以50万只普通绒鸭与大雪雁为例》。文件`COEI_data.csv`与`GSGO_data.csv`分别涵盖179幅COEI图像与99幅GSGO的ECA群图像的相关数据及估算结果,每张图像对应数据文件中的一行。两个文件均包含9列数据,对应变量说明如下:"Filenumber":用于标识对应图像的编号,大雪雁数据集的编号范围为"001"至"099",普通绒鸭数据集的编号范围为"001"至"179";"N":图像中已标注鸟类的总数量;"Nest":通过单次CountEm实模式运行得到的鸟类数量估算值(bN 439);"CE_sim":通过模拟模式下2000次模拟测量得到的经验误差系数(相对标准误差,CEe(bN 440));"f":CountEm实模式运行中采用的抽样比例;"n0":CountEm实模式运行中初始设置的样方数量;"Q":CountEm实模式估算过程中统计的鸟类总数量(即样本量);"n":CountEm实模式估算过程中统计的非空样方数量;"UserTime":CountEm实模式估算所用的计数时长(单位:分钟);"ManualTime":使用ImageJ完成手动标注流程所需的计数时长(单位:分钟)。

提供机构:
Zenodo
创建时间:
2022-01-08
二维码
社区交流群
二维码
科研交流群
商业服务