一種基于邊緣的圖像配準方法
An Image Edge Based Algorithm for Multisensor Image Registration
-
摘要: 圖像配準是多傳感器圖像融合研究中的一項關(guān)鍵技術(shù),多傳感器圖像特別是波段相距較遠的相關(guān)性較小的圖像,要實現(xiàn)圖像配準存在很大的困難。對于能夠用仿射變換模型來描述的圖像,圖像之間比較明顯的特征是各個物體之間的邊緣。該文研究利用小波變換的方法提取圖像的邊緣,并對邊緣圖像作交互方差分析,搜索出最佳交互方差的配準參數(shù)。通過對SPOT和TM圖像的處理,達到了精度較高的配準效果。
-
關(guān)鍵詞:
- 圖像配準; 圖像融合; 邊緣檢測
Abstract: Image registration is an very important technology in multisensor image fusion. There is little correlation among multisensor images, especially images with different bands. It is very difficult to matching these images. Among some images which can be described in affine model, edge is a very obvious character. In this article, wavelet analysis is used to extract image edge, then correlative deviation is calculated. When correlative deviation reaches the minor, the image registration parameters are obtained. The processing of SPOT and TM images demonstrates the efficacy of the proposed method. -
Brown L G. A survey of inage registration techniques[J].ACM Computing Surveys.1992, 24(4):325-376[2]Hui Li, Manjunath B S. A contour-based approach to multisensor image registration. IEEE Trans. on Image Processing, 1995, IP-4(3): 320-334.[3]Jun-Wei Hsieh. Image registration using a new edge-based approach[J].Computer Vision and Image Understanding.1997, 67(2):112-130[4]Mallat S. Singularity detection and processing with wavelets[J]. IEEE Trans. on Information Theory, 1992, IT-38(2): 617-643.[5]Mallat S. Characterization of signals from multiscale edges[J].IEEE Trans. on Pattern Analysis Machine Intelligence.1992, 14(7):710-732[6]王東峰.多模態(tài)圖像配準和大型圖像配準.[博士論文],北京:中國科學院電子學研究所,2002. -
計量
- 文章訪問數(shù): 2802
- HTML全文瀏覽量: 161
- PDF下載量: 1343
- 被引次數(shù): 0
下載:
