伪装效果评价模型的图像纹理特征提取方法  

Texture Feature Extraction Method of Camouflage Effect Evaluation Model

作　　者：见超超 胡江华[1] 崔光振 

机构地区：[1]解放军理工大学电磁环境效应与电光工程重点实验室,江苏南京210007

出　　处：《指挥控制与仿真》2017年第3期102-105,共4页Command Control & Simulation

摘　　要：对伪装后的军事目标进行伪装效果评价,通常依据伪装前后目标图像特征的改变情况进行评价。光照环境影响导致采集的图像质量下降,对提取目标图像特征会产生不利影响,从而影响伪装效果的评价。针对光照环境对目标图像特征的影响,提出一种结合局部二进制模式(LBP)与灰度共生矩阵的目标纹理特征提取方法。该方法首先通过局部二进制模式(LBP)算法提取目标图像不受光照影响的图像特征信息,然后对局部二进制模式(LBP)图像提取灰度共生矩阵,获得描述图像纹理特征的5个特征值。通过对比实验表明该方法可以有效降低光照环境因素对伪装效果评价的不利影响,使伪装目标的伪装效果评价结果更加准确可靠。To evaluate camouflage effect of the camouflaged militar targets,we usually base on the changes of image char-acteristics before and after for camouflage target. Because of the influence of illumination environment led to the decrease of the image quality,it can produce adverse effect on the image feature and affect the camouflage effect evaluation. A method of Image texture feature extraction ＂which combination of local binary pattern （LBP） and gray level co-occurrence matrix is pro-posed in this paper. This method first by local binary pattern （LBP） algorithm to extract the image feature which is not af-fected by light inlor^ nation,then get the five characteristics values of image texture feature by gray level co-occurrence matrix. Through comparative experiments show that this method can effectively reduce the illumination of the adverse impact of envi-ronmental factors on the camouflage effect evaluation, camouflage effect evaluation result more accurate and reliable.

关 键 词：图像纹理特征 光照环境影响 局部二进制模式 灰度共生矩阵 纹理相似度 伪装效果评价 

分 类 号：E951.4[军事—军事工程] TP391[自动化与计算机技术—计算机应用技术]