窑开发与创新窑0引言在现代战争野发现即摧毁冶特点背景下袁伪装在军事应用中的地位不断提高遥迷彩伪装作为应用最广泛的伪装技术袁具有能够使目标与背景图像更好地融合尧更加有效对抗和迷盲高分辨率侦察设备和精确制导武器等多方面优势特点[1]遥迷彩的伪装效果评估是迷彩伪装设计与实施的关键环节遥传统的评估方法主要是基于观察员在野外条件下袁在一定距离上对伪装目标进行观察和识别袁评估结果较依赖于观察者的经验及主观判断袁客观性不强袁可重复性不高[2]遥目前现有迷彩伪装效果量化评估方法大多根据亮度对比尧色差等单一指标进行对比评估袁对战场目标伪装状态的反映不够全面袁缺乏较为科学的多指标伪装效果评估方法遥因此袁建立一套客观性强尧全面反映迷彩伪装效果的多指标量化评估方法是我军伪装学科亟待解决的问题遥本文基于数字图像特征提取方法袁构建迷彩伪装效果评估指标体系袁运用超效率数据包络分析渊DataEnvelop鄄mentAnalysis冤和灰色关联分析渊GreyCorrelationAnalysis冤方法袁研究建立一种更客观尧准确尧具有普适性的迷彩伪AnCamouflageEffectEvaluationMethodofPatternPaintingBasedonSuper-efficientDEA-GCAZHANGXiao-Kun1袁2袁LILing1袁WANGXin-Xin3袁LIJiang2渊1.ArmyEngineeringUniversityofPLACollegeofFieldEngineering袁NanjingJiangsu210000袁China曰2.Unit91112ofPLA袁NingboZhejiang315000袁China曰3.AirForceLogisticsUniversity袁XuzhouJiangsu221000袁China冤Abstract院Aimingattheproblemsoftraditionalpatternpaintingcamouflageeffectevaluation袁suchasstrongsubjectivityandsingleindex袁thispaperproposesanevaluationmethodbasedonsuper-efficiencydataenvelopmentanalysis渊DEA冤andgreycorrelationanalysis渊GCA援Constructapatternpaintingcamouflageevaluationindexsystemincludingcomprehensiveimagefeaturesofstatistics袁color袁texture袁shape袁andcomplexity袁andanimagesimilaritymodelbasedonGaussianManhattandistance援Thesuper-efficiencyDEA-GCAhybridmodelisestablished袁theimagesimilarityisusedastheDEAoutputmatrix袁andthesuper-efficiencyvalueisusedasthegreycorrelationreferencesequence援Then袁thecamouflageeffectsof2patternpaintingsunder8typicalbackgroundsareexperimentallyverified援Itcanbeseenthatthismethodcaneffectivelyquantifyandevaluatethecamouflageeffectofpatte...
