计算机工程与设计
計算機工程與設計
계산궤공정여설계
COMPUTER ENGINEERING AND DESIGN
2010年
7期
1533-1535,1539
,共4页
图像融合%相关系数%融合规则%平均梯度%邻域%双数复小波变换
圖像融閤%相關繫數%融閤規則%平均梯度%鄰域%雙數複小波變換
도상융합%상관계수%융합규칙%평균제도%린역%쌍수복소파변환
image fusion%correlation coefficient%fusion rules%average gradient%neighborhood%dual-tree complex wavelet transform
通过分析研究图像融合的客观评价指标,提出了一种新的基于邻域的融合策略,即首先对源图像进行双树复小波(DT-CWT)分解,得到高频和低频图像.针对低频图像采用相关系数为阚值,以标准差加权平均进行邻域融合,高频图像以平均梯度为测度参数进行邻域融合,最后进行逆变换得到融合图像.采用均值、方差、熵和平均梯度4种客观评价指标来评价融合图像效果.实验结果表明,该方法能够较好地适应相似度差异较大的多传感器图像融合,有效地增强融合图像的细节信息和清晰度.
通過分析研究圖像融閤的客觀評價指標,提齣瞭一種新的基于鄰域的融閤策略,即首先對源圖像進行雙樹複小波(DT-CWT)分解,得到高頻和低頻圖像.針對低頻圖像採用相關繫數為闞值,以標準差加權平均進行鄰域融閤,高頻圖像以平均梯度為測度參數進行鄰域融閤,最後進行逆變換得到融閤圖像.採用均值、方差、熵和平均梯度4種客觀評價指標來評價融閤圖像效果.實驗結果錶明,該方法能夠較好地適應相似度差異較大的多傳感器圖像融閤,有效地增彊融閤圖像的細節信息和清晰度.
통과분석연구도상융합적객관평개지표,제출료일충신적기우린역적융합책략,즉수선대원도상진행쌍수복소파(DT-CWT)분해,득도고빈화저빈도상.침대저빈도상채용상관계수위감치,이표준차가권평균진행린역융합,고빈도상이평균제도위측도삼수진행린역융합,최후진행역변환득도융합도상.채용균치、방차、적화평균제도4충객관평개지표래평개융합도상효과.실험결과표명,해방법능구교호지괄응상사도차이교대적다전감기도상융합,유효지증강융합도상적세절신식화청석도.
By analyzing object evaluation of image fusion, a new method for image fusion based on neighborhood is proposed. Firstly, dual-tree complex wavelet is applied into source images to get high-frequency and low-frequency images. Secondly, weighted average neighborhood-standard-deviation combined with correlation coefficient threshold is adopted on low-frequency images. Neighborhood-Average-Gradient is adopted on high-frequency. Finally, the inverse transform is utilized to obtain fused image. The performance of the image fusion method is evaluated using four criteria including average value, square difference, entropy and average grads. Experimental results show that the proposed methods could adapt for the similarity of different multi-sensor image fusion and better for detail and definition.
