NWPU VHR-10

NWPU VHR-10数据集是具有挑战性的十类地理空间对象检测数据集。该数据集总共包含800幅VHR光学遥感图像，其中715幅彩色图像是从谷歌地球获得的，空间分辨率从0.5到2米，从具有0.08 m的空间分辨率的Vaihingen数据获取85幅经锐化的彩色红外图像。该数据集被分成两组: a) 在图像中包含至少一个目标的正图像集包含650图像。b) 负图像集包含150图像，并且它不包含任何目标。由此，正面图像集、757飞机、302船只、655储罐、390棒球钻石、524网球场、159篮球场、163地面田径场、224港口、124桥梁和477车辆被手动注释为边界框和用于地面真相的实例面具。

The NWPU VHR-10 dataset is a challenging ten-class geospatial object detection dataset. It contains a total of 800 very high-resolution (VHR) optical remote sensing images, among which 715 color images were acquired from Google Earth with a spatial resolution ranging from 0.5 to 2 meters, and 85 sharpened color infrared images were derived from the Vaihingen dataset which has a spatial resolution of 0.08 meters. The dataset is divided into two subsets: a) The positive image set, which contains at least one target object in each image, consists of 650 images; b) The negative image set, which contains no target objects whatsoever, consists of 150 images. Within the positive image set, a total of 757 airplanes, 302 ships, 655 storage tanks, 390 baseball diamonds, 524 tennis courts, 159 basketball courts, 163 ground track fields, 224 ports, 124 bridges, and 477 vehicles were manually annotated with bounding boxes and instance masks as ground truth.

搜集汇总

背景与挑战

背景概述

NWPU VHR-10是一个用于地理空间对象检测的遥感数据集，包含800幅高分辨率光学图像，分为650幅正样本（含十类目标如飞机、车辆等）和150幅负样本，所有目标均标注了边界框和实例掩码。该数据集具有多源图像特点（来自谷歌地球和Vaihingen数据），适用于计算机视觉和遥感分析任务。