航天返回与遥感
航天返迴與遙感
항천반회여요감
SPACECRAFT RECOVERY & REMOTE SENSING
2015年
2期
69-76
,共8页
多相位小波变换%影像复原%自适应融合%航天遥感
多相位小波變換%影像複原%自適應融閤%航天遙感
다상위소파변환%영상복원%자괄응융합%항천요감
multi-phase wavelet transformation%image restoration%self-adaption fusion%space remote sensing
IHS(intensity-hue-saturation)变换是影像融合实际生产中使用最多的一种方法,而小波变换是近几年来影像融合中的热门研究方向。但是现有方法存在纹理畸变和光谱畸变的现象,尤其是当地物光谱特性在全色和多光谱中存在较大差异的时候,融合后光谱畸变将会十分突出。为了解决上述问题,在分析现有小波变换方法的基础上提出了一种基于加权小波变换及调制传递函数补偿(MTFC)的多光谱影像融合方法,通过引入多相位小波变换的方式来抑制小波变换产生的纹理畸变,同时通过引入MTFC的方法来恢复影像融合中丢失的纹理信息。文章选用“高分二号”卫星影像来验证算法的有效性,试验结果表明,与现有的融合方法相比,文章中提出的算法能够很好地抑制影像中的光谱畸变,同时保留更多的有效纹理信息。
IHS(intensity-hue-saturation)變換是影像融閤實際生產中使用最多的一種方法,而小波變換是近幾年來影像融閤中的熱門研究方嚮。但是現有方法存在紋理畸變和光譜畸變的現象,尤其是噹地物光譜特性在全色和多光譜中存在較大差異的時候,融閤後光譜畸變將會十分突齣。為瞭解決上述問題,在分析現有小波變換方法的基礎上提齣瞭一種基于加權小波變換及調製傳遞函數補償(MTFC)的多光譜影像融閤方法,通過引入多相位小波變換的方式來抑製小波變換產生的紋理畸變,同時通過引入MTFC的方法來恢複影像融閤中丟失的紋理信息。文章選用“高分二號”衛星影像來驗證算法的有效性,試驗結果錶明,與現有的融閤方法相比,文章中提齣的算法能夠很好地抑製影像中的光譜畸變,同時保留更多的有效紋理信息。
IHS(intensity-hue-saturation)변환시영상융합실제생산중사용최다적일충방법,이소파변환시근궤년래영상융합중적열문연구방향。단시현유방법존재문리기변화광보기변적현상,우기시당지물광보특성재전색화다광보중존재교대차이적시후,융합후광보기변장회십분돌출。위료해결상술문제,재분석현유소파변환방법적기출상제출료일충기우가권소파변환급조제전체함수보상(MTFC)적다광보영상융합방법,통과인입다상위소파변환적방식래억제소파변환산생적문리기변,동시통과인입MTFC적방법래회복영상융합중주실적문리신식。문장선용“고분이호”위성영상래험증산법적유효성,시험결과표명,여현유적융합방법상비,문장중제출적산법능구흔호지억제영상중적광보기변,동시보류경다적유효문리신식。
AbstractImage fusion is an important tool to fuse high spectral and high spatial information into one image for image interpretation and target recognition in remote sensing application. To date, many image fusion methods have been developed, among which IHS technique is the most widely used, and the wavelet fusion is the most frequently discussed in recent publications due to its obvious advantages. However, the available methods can hardly produce a satisfactory fusion result, in which spectral and spatial distortions often take place, especially when spectral properties of the same surface features are different between natural color images and panchromatic ones. To solve the problem, we propose a new method by introducing a multi-phase approach to restrain spatial distortion caused by the shift variant attribution of wavelet transformation. With suitable rules of fusion, the method uses modulation transfer function compensation (MTFC) to restore details, so that the spectral distortion can be avoided. The performance of the method was evaluated with standard images from GF-2 satellite. It was proven that the method could improve the fusion quality significantly in comparison with conventional fusion methods.