参考文献:
国内从事红外热成像的公司
近红外相机高分辨成像将迎来行业爆发(一)
近红外相机高分辨成像将迎来行业爆发(二)
红外热像仪助力自动驾驶安全
“优势:FLIR的热成像传感器则主要利用远红外光谱。它们能够探测低至0.1华氏度的微小温差,因此,即使是在寒冬的夜间,纤细的自行车轮毂也能清晰呈现,而且,它的探测距离最大超过240米,这个探测距离跟目前市场上性能最强劲的少数几款LiDAR不相上下。此外,跟LiDAR系统不同,红外热成像传感器无惧浓雾或阳光直射。”
“劣势:红外热像仪的分辨率没有传统可见光摄像头高。而且,红外光波无法穿透透明的玻璃,因此,没有任何热像仪能够“看”到玻璃后方的潜在威胁”
“Seek Thermal在年初的CES 2018上便展出了其首款应用于汽车后装市场的高分辨率红外热成像摄像头,拥有320 x 240 像素高分辨率热成像传感器和24度视场角,搭载了双元素硫系镜头,能够非常方便地与现有车载娱乐系统集成,售价预计将低于999美元。”
基于红外热成像的行人检测方法 方法来自于《Thermal-Infrared Pedestrian ROI Extraction through Thermal andMotion Information Fusion》
(xys—基于二值分割)
基于卷积神经网络的近红外夜间道路行人识别 near infrared nighttime road pedestrians recognition based on convolutional neural network
深度学习方法实现红外图片中人物动作识别 deep learning approach for human action recognition in infrared images
红外动作识别数据集:不同季节红外动作识别的研究 InfAR dataset Infrared action recognition at different times
兄弟文章:A New Dataset and Evaluation for infrared action recognition(重庆邮电大学 高陈强)
图像融合论文及代码网址整理总结(2)——红外与可见光图像融合
红外和可见光图像融合_图像融合数据集
基于可见光、红外视频融合的夜间多目标实时检测算法
在弱光条件下使用辅助卷积神经网络和可见光图像的红外图像超分辨率
Lan Wang, Chenqiang Gao, Luyu Yang, Yue Zhao, Wangmeng Zuo,
and Deyu MengECCV 2018[ If you want to download the dataset,
please contact: gaocq@cqupt.edu.cn.
You are required to provide necessary information, including: your name, group, affiliation, usage purpose, etc.
If you have any questions on technical details, please contact: sofellsune@yeah.net ]
(The dataset is divided in two:
(i) Classification dataset: positives and randomly sampled negatives with a fixed height-width ratio of (1/2) and rescaled to 64x32 pixels,
(ii) Detection Dataset: Original positive and negative images with annotations.)
Terravic Weapon IR Database—红外武器数据库
OTCBVS—俄亥俄大学红外人体数据集
KAIST行人数据集
KAIST: Multispectral Pedestrian Detection Benchmark
对应的论文:cvpr2015-Multispectral Pedestrian Detection: Benchmark Dataset and Baseline
Description
We developed imaging hardware consisting of a color camera, a thermal camera and a beam splitter
to capture the aligned multispectral (RGB color + Thermal) images. With this hardware,
we captured various regular traffic scenes at day and night time to consider changes in light conditions.
The KAIST Multispectral Pedestrian Dataset consists of 95k color-thermal pairs (640x480, 20Hz) taken
from a vehicle. All the pairs are manually annotated (person, people, cyclist) for the total
of 103,128 dense annotations and 1,182 unique pedestrians. The annotation includes temporal
correspondence between bounding boxes like Caltech Pedestrian Dataset. More infomation can
be found in our CVPR 2015 paper.(Multispectral Pedestrian Detection: Benchmark Dataset and Baseline)
**KAIST行人数据集: 博客,介绍了该数据集上的清洗工作,以及数据集的github主页,以及在github上的下载链接, 还有另一个地址
** 关于KAIST数据集-(xys-无法打开图像问题的解决)
SCUT_FIR_Pedestrian_Dataset : 华南理工,2019,下载地址, 论文:Benchmarking a large-scale FIR dataset for on-road pedestrian detection
Cats: A color and thermal stereo bench marks l论文中提供了下载链接
CATS 2数据集,可见光+红外+深度 论文: CATS 2: Color And Thermal Stereo Scenes with Semantic Labels ,2019
论文:2019 RGB-T object tracking- Benchmark and baseline,下载地址, 安徽大学,兄弟论文:Fast RGB-T Tracking via Cross-Modal Correlation Filters