激光与红外
激光與紅外
격광여홍외
LASER & INFRARED
2014年
3期
325-329
,共5页
蒋立辉%高浩%庄子波%熊兴隆
蔣立輝%高浩%莊子波%熊興隆
장립휘%고호%장자파%웅흥륭
低空风切变%目标识别%组合纹理%激光雷达图像
低空風切變%目標識彆%組閤紋理%激光雷達圖像
저공풍절변%목표식별%조합문리%격광뢰체도상
low altitude wind shear%target recognition%combination texture%laser radar image
针对不同的风切变在激光雷达图像上所呈现的不同纹理特性,提出了一种组合局部纹理特征和全局纹理特征的识别方法。先分别从激光雷达风切变图像中提取 LBP 特征和灰度-梯度共生矩阵特征,LBP特征反应图像的局部纹理,代表风场局部风速的变化,灰度-梯度共生矩阵特征反应图像的全局纹理,代表风场全局的风速变化,再通过典型相关分析对两种特征进行融合,最后采用最近邻分类器对三种风切变进行匹配识别。实验结果表明,该算法对三种低空风切变的平均识别率达到99.02%,与三种单一的纹理特征分类识别相比,分别提高了18.86%,5.88%和7.01%。
針對不同的風切變在激光雷達圖像上所呈現的不同紋理特性,提齣瞭一種組閤跼部紋理特徵和全跼紋理特徵的識彆方法。先分彆從激光雷達風切變圖像中提取 LBP 特徵和灰度-梯度共生矩陣特徵,LBP特徵反應圖像的跼部紋理,代錶風場跼部風速的變化,灰度-梯度共生矩陣特徵反應圖像的全跼紋理,代錶風場全跼的風速變化,再通過典型相關分析對兩種特徵進行融閤,最後採用最近鄰分類器對三種風切變進行匹配識彆。實驗結果錶明,該算法對三種低空風切變的平均識彆率達到99.02%,與三種單一的紋理特徵分類識彆相比,分彆提高瞭18.86%,5.88%和7.01%。
침대불동적풍절변재격광뢰체도상상소정현적불동문리특성,제출료일충조합국부문리특정화전국문리특정적식별방법。선분별종격광뢰체풍절변도상중제취 LBP 특정화회도-제도공생구진특정,LBP특정반응도상적국부문리,대표풍장국부풍속적변화,회도-제도공생구진특정반응도상적전국문리,대표풍장전국적풍속변화,재통과전형상관분석대량충특정진행융합,최후채용최근린분류기대삼충풍절변진행필배식별。실험결과표명,해산법대삼충저공풍절변적평균식별솔체도99.02%,여삼충단일적문리특정분류식별상비,분별제고료18.86%,5.88%화7.01%。
As different wind shear presents different characteristics of texture on the laser radar images,a recognition algorithm which combines local and global texture features is proposed.Firstly,local Binary Pattern(LBP)features and Gray-Gradient Co-occurrence Matrix(GGCM)features are extracted from the laser radar images respectively,LBP features react the local texture of the images and represent the changes of wind speed of local wind farm,GGCMfea-tures react the global texture of the images and represent the speed changes of whole wind field.The two features are fused through Canonical Correlation Analysis(CCA),finally the nearest neighbor classifier is adopted to match three different wind shears.Experiment results show that the recognition rate of the proposed algorithm on the three kinds of low altitude wind shear can reach 99.02%.Compared with three kinds of single texture,the recognition rate is raised by 18.86%,5.88% and 7.01% respectively.