基于红外相机技术对祁连山区哺乳动物多样性的调查：兰州大学祁连山区寺大隆样区红外相机数据（2020）

该数据集包含了2020年1月至2020年10月的兰州大学祁连山区寺大隆样区红外相机数据。寺大隆样区的典型生境是森林，主要树种为祁连圆柏和青海云杉，典型的哺乳动物有马鹿、马麝、狍、蓝马鸡等。
红外相机数据处理的主要步骤包括：
1.数据存储，在计算机、移动硬盘或其他存储介质上设置目录存储照片及录像文件。
2.处理误拍或无效照片。删除风吹草动、曝光、没有动物存在或者任意形式的无效照片。3.物种鉴定。
（1）动物识别图像库建设，每一个调查单元建立一个动物识别图像库，该库主要用于物种识别人员的培训，便于其快速掌握物种鉴别特征，准确识别物种。
（2）有效照片的处理：对于能够准确识别物种的照片（录像），在自动相机（视频）记录表填写动物名称、数量、环境信息等；如果一张照片上有两种动物或两种以上动物，则各填写一行；对于不能准确识别物种的照片，在动物名称一栏填写无法识别，并填写数量、环境信息等，并在照片处理一栏填写“未处理”；对于家禽家畜等，填写动物家禽家畜的名称及数量；对于人，在动物名称处填写“牧民、游客、护林员”等。
（3）其它信息：环境信息记录，应根据照片（录像），填写以下环境信息：温度：依照照片上显示的温度填写。天气：晴、阴、雨、雪。需要仔细判断。积雪：有或无。行为：觅食、饮水、捕食、交配、打斗、争食、驱赶、嬉戏、奔跑、休息、行走、警戒等。动物年龄：幼、亚、雌、雄、未知。
发布的观测数据包括：文件编号，文件格式，文件夹编号，相机编号，布设点位编号，拍摄日期，拍摄时间，工作天数（天），要素，物种名称，幼，亚，雌，雄，未知，总数，行为，温度（℃），天气，积雪。

This dataset contains infrared camera trap data collected from the Sidalong sampling area in the Qilian Mountains of Lanzhou University, spanning from January to October 2020. The typical habitat of the Sidalong sampling area is forest, with the dominant tree species being *Sabina przewalskii* and *Picea crassifolia*, and common mammalian and avian species including Red Deer (*Cervus elaphus*), Alpine Musk Deer (*Moschus chrysogaster*), Siberian Roe Deer (*Capreolus pygargus*), and Blue Eared-Pheasant (*Crossoptilon auritum*), among others.

The main processing steps for the infrared camera trap data are as follows:
1. Data Storage: Establish directories on computers, external hard drives or other storage media to store captured photos and video files.
2. Removal of False-Triggered and Invalid Photos: Delete photos caused by wind-blown vegetation, overexposure, absence of target animals, or any other form of invalid content.
3. Species Identification:
(1) Construction of Animal Identification Image Libraries: Build an animal identification image library for each survey unit, which is primarily used to train species identifiers to quickly grasp species identification traits and accurately recognize animal species.
(2) Processing of Valid Photos/Videos: For photos (videos) that allow accurate species identification, fill in the animal name, quantity, environmental information and other details in the camera trap record sheet. If a single photo contains two or more animal species, fill in one separate row for each species. For photos that cannot be accurately identified to species level, fill in "unidentifiable" in the animal name column, along with the quantity and environmental information, and mark "unprocessed" in the photo processing column. For domestic poultry and livestock, fill in their respective names and quantities. For humans, fill in terms such as "herdsmen, tourists, forest rangers" in the animal name column.
(3) Supplementary Environmental Information: Record environmental information based on the captured photos or videos as follows:
- Temperature: Fill in the temperature value displayed on the photo.
- Weather: Sunny, cloudy, rainy, snowy (requires careful visual judgment).
- Snow Cover: Present or Absent.
- Behavior: Foraging, drinking water, predation, mating, fighting, competing for food, displacing conspecifics, playing, running, resting, walking, alert vigilance, etc.
- Animal Age: Juvenile, Sub-adult, Female, Male, Unknown.

The published observation data includes the following fields: file number, file format, folder number, camera ID, survey station number, shooting date, shooting time, working days (unit: day), elements, species name, juvenile count, sub-adult count, female count, male count, unknown age count, total count, behavior, temperature (℃), weather, snow cover.

搜集汇总

背景与挑战

背景概述

该数据集是兰州大学在2020年1月至10月期间，使用红外相机技术在祁连山区寺大隆样区采集的哺乳动物多样性观测数据。数据集覆盖森林生境，主要记录马鹿、马麝、狍、蓝马鸡等物种，并包含物种行为、环境温度、天气等详细观测信息，数据以申请获取方式共享，适用于生态研究和生物多样性监测。