夜视环境下的雷达遥感图像的小特征分割仿真  

Small Feature Segmentation Simulation of Radar Remote Sensing Images under Night Vision Environment

作　　者：唐昌华[1] 庄庆华[1] 

机构地区：[1]长春工业大学人文信息学院,吉林长春130000

出　　处：《计算机仿真》2015年第12期22-25,共4页Computer Simulation

基　　金：吉林省教育厅"十二五"科学技术研究项目(吉教科合字[2014]第641号)

摘　　要：在雷达遥感图像质量优化中,由于数据在采集、传输、处理的各个环节出现高斯白噪声干扰时,采用传统的方法,使得遥感图像多区域梯度直方分割受到这种干扰的影响,图像信息数据采集不准确,导致夜视环境下的远程雷达遥感图像不完整,识别效果差的问题。提出一种夜视环境下的远程雷达遥感图像的小特征分割算法,用双树小特征分割法对遥感图像多区域梯度直方分割,按级别对遥感图像信息数据进行采集,数据逆变换后进行扩散图像去噪,保持夜视环境下的远程雷达遥感成像完整性。仿真结果证明,小特征分割方法比原有方法的综合性能大大提高,改善了原有的雷达夜间成像不完整,成像效果差,识别判断难度大等缺点。改进算法解决了弊端,提高了夜视环境下的远程雷达遥感图像的效果。In the paper, a small feature segmentation algorithm for long - range radar remote sensing image under night vision environment was proposed. A binary tree small feature segmentation method was used to make multi - re- gional gradient histogram segmentation of remote sensing image. According to the level, information data of remote sensing image were collected. Diffusion image denoising was made after data inverse transformation to maintain the integrity of long - range radar remote sensing image under the night vision environment. The simulation results show that the comprehensive performance of the small feature segmentation method is greatly improved compared with the original method, to perfect the shortcomings of the incomplete original radar imaging in night, the poor imaging effect, and the difficult identifying and judging. The improved algorithm has solved the problem and improved the effect of the long - range radar remote sensing image under the night vision environment.

关 键 词：夜视环境 远程雷达 小特征分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]