five

Panama Vegetation Time Series Maps|植被监测数据集|森林砍伐数据集

收藏
Mendeley Data2024-01-31 更新2024-06-27 收录
植被监测
森林砍伐
下载链接:
https://figshare.com/articles/dataset/Panama_Vegetation_Time_Series_Maps/14120603
下载链接
链接失效反馈
资源简介:
Panama Vegetation Time Series Maps for 1990-2016 with preliminary 2020 deforestation data. (the official 2020/21 map will be made available at the end of the 2021 mapping period). The following components are provided: "PVCTSYYYYvN" are 30m resolution raster maps resulting from compositing of 60-100 classified Landsat images for each time step, where YYYY is the nominal year of the compositing time period. (1991 includes images from 1987-1991. 2001 includes images from 1997-2001. 2006 includes images from 2002-2006. 2011 includes images from 2007-2011. 2016 includes images from 2012-2016.) Maps are all in GeoTiff format. "PVCTS_key.lyr" is a layer file that can be applied in ArcGIS for optimal display of categories. "PVCTSColorKey" provides more detailed description of categories and can be used to create a key in other software. More details about the land cover classes can be found in the "PVCTSv2_SupplementaryInfo" file. "PVCTSdeforest7cat_YYXX" are corresponding deforestation maps for activity occurring between years YY and XX. "Deforest7cat.lyr" provides the layer file that can be applied in ArcGIS for suggested viewing and "PVCTS_Deforestation_ColorKey" provides a description of the categories and can be used to create a key in other software. "PVCTSv2_SupplementaryInfo" provides information about the methods and data in the PVCTS composite and deforestation files. Accuracy assessments and error adjustments for the current PVCTS version are included in this document. "MaxAgeYYYY" are maximum-vegetation-age raster maps based on aggregation of clearing observations from all images in the USGS Landsat archive with cloud-cover <70%. These are 30m resolution rasters in GeoTiff format. Values indicate the maximum age that the vegetation can be in year YYYY based on the most recent clearing observation, or 99 if never observed to be cleared. A value of -1 means that the pixel was not observed to be cleared in year YYYY but it was observed to be cleared in year YYYY+1. A clearing observation occurs when the NBR (Normalized Burn Index) is below a season-based threshold (1000 to 3000) in a cloud-masked image. Many pixels are not observed each year due to cloud cover, thus these figures represent Maximum age (vegetation might be younger). The information in these layers has already been applied to the PVCTS maps, but not vice-versa (no other data is taken into account beyond NBR threshold). "metadata_spatial_ss" provides the spatial metadata for all GeoTiff files. Projection is in WGS_1984_UTM_Zone_17N.
创建时间:
2024-01-31
用户留言
有没有相关的论文或文献参考?
这个数据集是基于什么背景创建的?
数据集的作者是谁?
能帮我联系到这个数据集的作者吗?
这个数据集如何下载?
点击留言
数据主题
具身智能
数据集  4098个
机构  8个
大模型
数据集  439个
机构  10个
无人机
数据集  37个
机构  6个
指令微调
数据集  36个
机构  6个
蛋白质结构
数据集  50个
机构  8个
空间智能
数据集  21个
机构  5个
5,000+
优质数据集
54 个
任务类型
进入经典数据集
热门数据集

Breast Ultrasound Images (BUSI)

小型(约500×500像素)超声图像,适用于良性和恶性病变的分类和分割任务。

github 收录

MultiTalk

MultiTalk数据集是由韩国科学技术院创建,包含超过420小时的2D视频,涵盖20种不同语言,旨在解决多语言环境下3D说话头生成的问题。该数据集通过自动化管道从YouTube收集,每段视频都配有语言标签和伪转录,部分视频还包含伪3D网格顶点。数据集的创建过程包括视频收集、主动说话者验证和正面人脸验证,确保数据质量。MultiTalk数据集的应用领域主要集中在提升多语言3D说话头生成的准确性和表现力,通过引入语言特定风格嵌入,使模型能够捕捉每种语言独特的嘴部运动。

arXiv 收录

WenetSpeech

WenetSpeech是一个大规模的多领域普通话语音数据集,总时长超过22400小时,包括10000+小时高质量标注语音、2400+小时弱标注语音和约10000小时未标注语音。数据来源于YouTube和Podcast,覆盖多种说话风格、场景、领域、话题和噪声条件。通过光学字符识别(OCR)方法和高质量自动语音识别(ASR)转录系统生成音频/文本分割候选,并采用端到端标签错误检测方法进行验证和过滤。此外,提供三个手动标注的高质量测试集用于评估。WenetSpeech是目前最大的开源普通话语音语料库,适用于生产级语音识别研究。

arXiv 收录

flames-and-smoke-datasets

该仓库总结了多个公开的火焰和烟雾数据集,包括DFS、D-Fire dataset、FASDD、FLAME、BoWFire、VisiFire、fire-smoke-detect-yolov4、Forest Fire等数据集。每个数据集都有详细的描述,包括数据来源、图像数量、标注信息等。

github 收录

MMRS - 多模态遥感指令跟随数据集

MMRS数据集,包含了100万多个图像-文本对,涵盖了分类、检测、图像描述、VQA、视觉定位等多个任务,并包括光学、红外和SAR三种视觉模态。该数据集旨在促进遥感领域中MLLMs的持续发展。

arXiv 收录