摘要
为了给用户提供更为直观可靠的数据挖掘结果,提高数据处理精度及效率,设计了基于激光内调制技术的高光谱数据可视化挖掘系统。首先使用激光内调制技术控制激光形成,加强激光雷达调制效率,运用基于VMD的回波信号去除图像噪声;其次构建高光谱彩色可视化模型,把评估数据降采样处理,利用像元之间彩色差别均值保证图像可分性,通过局部优化手段修改图像,实现图像可视化呈现;最后通过ENVI平台和IDL语言建立高光谱数据可视化挖掘系统,包含数据输入输出、数据重采样、数据预处理和待测目标处理4个板块。仿真实验结果表明,系统可以快速准确挖掘目标高光谱数据,观察到数据内隐藏的特征信息数据,为光谱数据分析提供有效帮助。
In order to provide users with more intuitive and reliable data mining results,and to improve the accuracy and efficiency of data processing,this paper designs a hyperspectral data visualization mining system based on laser internal modulation technology.Firstly,we use laser internal modulation technology to control the laser formation,strengthen the modulation efficiency of lidar,and use the echo signal based on VMD to remove image noise.Secondly,we build a hyperspectral color visualization model to reduce the sample processing of evaluation data,use the mean value of color difference among pixels to ensure the image separability,modify the image through local optimization,and realize the image visualization.After that,the hyperspectral data visualization mining system is established by ENVI platform and IDL language,including data input and output,data resampling,data preprocessing and target processing.The simulation results show that the system can mine hyperspectral data quickly and accurately,observe the hidden feature information data,and provide effective help for spectral data analysis.
作者
胡洋
肖剑
张娅妮
HU Yang;XIAO Jian;ZHANG Yani(Qiannan Normal University for Nationalities,Duyun Guizhou 558000,China)
出处
《激光杂志》
北大核心
2020年第10期167-171,共5页
Laser Journal
基金
贵州省科技厅联合基金项目(No.黔科合LH字[2014]7440)。
关键词
内调制技术
高光谱图像
可视化挖掘系统
图像去噪
internal modulation technology
hyperspectral image
visual mining system
image denoising
作者简介
胡洋(1970-),男,副教授,研究方向:数据挖掘、数据可视化系统开发。
