多传感器多尺度图像信息融合算法-文成林.pdfVIP

5 　　　　　　 Vol.36 　No.5 　2008 5 ACTA ELECTRON CA S N CA May 　2008 　 1 2 2 文成林, 郭　超, 高敬礼 (1., 310018;2., 475001) 　　　:　, .:, , ;, , ; ;, ., , . :　;;; :　TP391　　　:　A 　　　:　0372-2112 (2008)05-0840-08 Multiscale Image Informat ion from Multisensor Fusion Algorit hm 1 2 2 WEN Cheng-lin , GUO Chao ,GAO Jing-li (1.Institute of Inf ormation and Automation, Hangzhou Dianzi University, Hangzhou, Zhej iang 310018, China; 2.School of Computer and Inf ormation Engineering, Henan University, Kaif eng, Henan 475001, China) Abstract:　While there are multiple images obtained by different sensors to measure the same object scene, one new multi- scale image fusion algorithm based on probabilistic model is proposed.Its elementary idea is:Firstly, to decompose each sensor im- age into multiple subimages which compose a multiscale pyramid via wavelet packet transform, and to establish pixel-based subim- age model on every level in the pyramid.Secondly, estimate the model parameters using least squares method based on the corre- sponding pixel of each level ofthe sensor image.Then, based on themodel, derive estimation of the true sceneusing maximum pos- terior method.Finally, we may obtain global fusion estimate with the object sceneby applying orderly inverse wavelet packet trans- formation to every local fusion estimatefrom each level of the pyramid.The result of fusing the visible image and infrared image shows that the proposed algorithm is valid. Key words:　wavelet packet transform;probabilistic model;least squares;maximam a pos