应用科技
應用科技
응용과기
YING YONG KE JI
2014年
5期
48-52
,共5页
图像融合%非采样Contourlet变换%灰色系统理论%目标提取%融合规则%红外线%可见光
圖像融閤%非採樣Contourlet變換%灰色繫統理論%目標提取%融閤規則%紅外線%可見光
도상융합%비채양Contourlet변환%회색계통이론%목표제취%융합규칙%홍외선%가견광
image fusion%nonsubsampled Contourlet transform%grey system theory%target extraction%fusion rule%infrared%visible light
针对灰色系统理论在空间域对红外与可见光图像融合的不足,以及非下采样Contourlet变换( NSCT)在图像融合领域的优势,提出了一种基于目标提取的红外与可见光图像融合新算法。首先,对红外和可见光图像分别进行NSCT变换;其次,对红外低频分量应用灰色系统理论进行目标提取,并利用所提融合规则对低频分量进行融合,同时对高频分量采用常用融合规则进行融合;最后,对融合后高、低频分量进行NSCT逆变换,得到融合图像。通过与4种常用方法进行实验对比,结果表明,文中算法得到的融合图像视觉效果较好,某些客观评价指标提升明显。
針對灰色繫統理論在空間域對紅外與可見光圖像融閤的不足,以及非下採樣Contourlet變換( NSCT)在圖像融閤領域的優勢,提齣瞭一種基于目標提取的紅外與可見光圖像融閤新算法。首先,對紅外和可見光圖像分彆進行NSCT變換;其次,對紅外低頻分量應用灰色繫統理論進行目標提取,併利用所提融閤規則對低頻分量進行融閤,同時對高頻分量採用常用融閤規則進行融閤;最後,對融閤後高、低頻分量進行NSCT逆變換,得到融閤圖像。通過與4種常用方法進行實驗對比,結果錶明,文中算法得到的融閤圖像視覺效果較好,某些客觀評價指標提升明顯。
침대회색계통이론재공간역대홍외여가견광도상융합적불족,이급비하채양Contourlet변환( NSCT)재도상융합영역적우세,제출료일충기우목표제취적홍외여가견광도상융합신산법。수선,대홍외화가견광도상분별진행NSCT변환;기차,대홍외저빈분량응용회색계통이론진행목표제취,병이용소제융합규칙대저빈분량진행융합,동시대고빈분량채용상용융합규칙진행융합;최후,대융합후고、저빈분량진행NSCT역변환,득도융합도상。통과여4충상용방법진행실험대비,결과표명,문중산법득도적융합도상시각효과교호,모사객관평개지표제승명현。
Considering the shortcomings of the infrared and visible light image fusion based on grey system theory in the spatial domain, and utilizing the advantages of nonsubsampled Contourlet transform ( NSCT) in image fusion, a fusion algorithm of the infrared and visible light image based on target extraction is proposed.Firstly, the nonsub-sampled Contourlet transform is performed respectively on infrared image and visible light image.Secondly, the grey system theory is applied to the low frequency component of the infrared image for target extraction, and then the low frequency components are fused by making use of the proposed fusion rule.At the same time the common fusion rule is applied to the high frequency components.Finally, the reverse nonsubsampled Contourlet transform is performed on the fused low frequency part and high frequency part in order to obtain the fusion image.Compared with four commonly used methods, the results show that the fusion image has a good visual effect, and some objective evalua-tion indexes are improved obviously.