北京生物醫(yī)學(xué)工程

基于非冗余平移不變小波變換的醫(yī)學(xué)圖像配準(zhǔn)

Medical Image Registration Using the Nearly Shift-Insensitive and nonredundancy Discrete Wavelet Transform

作者：石宏理羅述謙

單位：首都醫(yī)科大學(xué)生物醫(yī)學(xué)工程學(xué)院（北京100069）

關(guān)鍵詞：圖像配準(zhǔn)；離散小波變換（DWT）；平移敏感性；數(shù)字濾波器

分類號(hào)：

出版年·卷·期（頁(yè)碼）：2010·29·6（594-598）

摘要：

配準(zhǔn)是圖像處理中一個(gè)非常重要的過(guò)程，但其運(yùn)算量通常非常大。離散小波變換（discrete wavelet transform，DWT）已成為減小運(yùn)算量的有效工具，然而，它對(duì)平移的敏感意味著原圖像特征的微小位移在其小波分解子圖中可能產(chǎn)生不可預(yù)料的變化。本文根據(jù)分析認(rèn)為平移敏感性是由DWT中降采樣過(guò)程引入的混疊造成的，并以此提出了設(shè)計(jì)小波濾波器的新方法。該方法通過(guò)減小混疊影響，從而減小了DWT的平移敏感性。設(shè)計(jì)結(jié)果表明此方法使DWT的平移敏感性得到了有效抑制，同時(shí)保持了非冗余性。最后，兩個(gè)醫(yī)學(xué)圖像的配準(zhǔn)過(guò)程表明了該小波的有效性和設(shè)計(jì)方法的合理性。

Discrete wavelet transform (DWT) has become an attractive tool of image registration，however，it is shift-sensitive.Some modified schemes of DWT，such as DTCWT (dual-tree complex wavelet transform)，always lead to high computational complexity in registration process because of redundancy.In this paper，it showed that the shift-sensitivity was caused by the downsampling operation in the analysis process，which could be expressed by the aliasing terms in the frequency-domain.A new scheme for the design of wavelets was proposed to approximately eliminate the effect of downsampling while remains the wavelet representation nonredundancy.The design example illustrated the shift sensitivity of wavelet representations was reduced efficiently without any redundancy.The registration results of two medical images further proved the validity of design scheme and effectiveness of the proposed wavelet.

參考文獻(xiàn)：

［1］田捷，包尚聯(lián)，周明全.醫(yī)學(xué)影像處理與分析［M］.北京:電子工業(yè)出版社，2003.

［2］Duncan J，Ayache N.Medical image analysis: progress over two decades and the challenges ahead［J］.IEEE Trans.on Pattern Analysis and Machine Intelligence，2000，22(1): 85-106.
［3］Lo CH.Medical image registration by maximization of mutual information［D］.US:Kent State University，2003.
［4］Brown L.A survey of image registration techniques［J］.ACM Comput Surv，1992，24(4): 325-376.
［5］Stone HS，Moigne JL，McGuire M.The translation sensitivity of wavelet-based registration［J］.IEEE Trans Pattern Anal.Machine Intell，1999，21:1074-1081.
［6］Allen RL，Kamangar FA，Stokely EM.Laplacian and orthogonal wavelet pyramid decompositions in coarse-to-fine registration［J］.IEEE Trans Signal Processing，1993，41:3536-3541.
［7］Cole-Rhodes A，Johnson KL，Moigne JL，et al.Multiresolution registration of remote sensing imagery by optimization of mutual information using a stochastic gradient［J］.IEEE Trans.Image Processing，2003，12(12):1495-1511.
［8］Zavorin I,Moigne JL.Use of Multiresolution Wavelet Feature Pyramids for Automatic Registration of Multisensor Imagery［J］.IEEE Trans.Image Processing，2005，14(6):770-782.
［9］Kingsbury BN.Image processing with complex wavelets［M］.London:Phil Trans Roy Soc.1999:2543-2560.
［10］Kingsbury BN.Complex wavelets for shift invariant analysis and filtering of signals［J］.Appl Comput Harmonic Anal，2000，10(3):234-253.
［11］Cohen A，Daubechies I,Feauveau JC.Biorthogonal bases of compactly supported wavelets［J］.Commun Pure Appl Mathematics，1992，45:485-560.
［12］Lazaridis G,Petrou M.Image Registration Using the Walsh Transform［J］.IEEE Trans.Image Processing，2006，15(8): 2343-2357.

服務(wù)與反饋：

【文章下載】【加入收藏】

提示：您還未登錄，請(qǐng)登錄！點(diǎn)此登錄

51黑料吃瓜在线观看,51黑料官网|51黑料捷克街头搭讪_51黑料入口最新视频