红外可见光融合数据集:M3DF
红外可见光融合数据集:M3DF
M3DF是一个红外可见光融合数据集,来自2022年CVPR论文。该数据集包含8400张图像,涵盖了大连理工大学校园、大连金石海滩国家旅游度假区以及大连市金州区主要道路等多个场景。数据集中的图像经过严格配准处理,可用于融合、检测和基于融合的检测等多种任务。
红外可见光融合数据集:M3DF
资料
说明:本篇数据集是Target-aware Dual Adversarial Learning and a Multi-scenario Multi-Modality Benchmark to Fuse Infrared and Visible for Object Detection中提出的,同样包含了目标检测部分。这个文章来自 2022 年 CVPR。
官方 github:https://github.com/dlut-dimt/TarDAL(很全面)
Google Drive:https://drive.google.com/drive/folders/1H-oO7bgRuVFYDcMGvxstT1nmy0WF_Y_6
Baidu Yun:https://pan.baidu.com/s/1GoJrrl_mn2HNQVDSUdPCrw?pwd=M3FD
官方 github 翻译
数据集预览
- Sensor: A synchronized system containing one binocular optical camera and one binocular infrared sensor. More details are available in the paper.
- 传感器:由一个双目光学摄像机和一个双目红外传感器组成的同步系统。更多的细节可在文件中找到。
Main scene: 主场景:
Campus of Dalian University of Technology.
大连理工大学校园。
State Tourism Holiday Resort at the Golden Stone Beach in Dalian, China.
国家旅游度假区在大连的金石海滩,中国。
Main roads in Jinzhou District, Dalian, China.
大连市金州区主要道路。
Total number of images: 图像总数:
8400 (for fusion, detection and fused-based detection)
8400(用于融合、检测和基于融合的检测)
600 (independent scene for fusion)
600(独立场景融合)
Total number of image pairs:
图像对总数:
4200 (for fusion, detection and fused-based detection)
4200(用于融合、检测和基于融合的检测)
300 (independent scene for fusion)
300(独立场景融合)
Format of images: 图像格式:
[Infrared] 24-bit grayscale bitmap
[红外] 24位灰度位图
[Visible] 24-bit color bitmap
[可见] 24位彩色位图
Image size: 1024 x 768 pixels (mostly)
图像大小:1024 x 768像素(大部分)
Registration: All image pairs are registered. The visible images are calibrated by using the internal parameters of our synchronized system, and the infrared images are artificially distorted by homography matrix.
配准:配准所有图像对。利用同步系统的内部参数对可见光图像进行了标定,利用单应矩阵对红外图像进行了畸变处理。
Labeling: 34407 labels have been manually labeled, containing 6 kinds of targets: {People, Car, Bus, Motorcycle, Lamp, Truck}. (Limited by manpower, some targets may be mismarked or missed. We would appreciate if you would point out wrong or missing labels to help us improve the dataset)
标注:已人工标注34407个标签,包含6种目标:{人、汽车、公交车、摩托车、灯具、卡车}。(受人手所限,可能会误标或漏标。如果您能指出错误或缺失的标签以帮助我们改进数据集,我们将不胜感激)