Accurate Image Registration Using SIFT for Extended-Field-of-View Sonography

摘要：

Extended-field-of-view (EFOV) sonography has beenproven to be capable of illustrating panoramic ultrasound images.In EFOV technology, accurate registration of original images is the basis. Conventional registration methods based on blockmatching cannot provide adequate accuracy when the relativerotations among the images are large. To register the images withimproved accuracy, this paper proposes a novel methodologicalframework based on SIFT (Scale Invariant Feature Transform)and Levenberg-Marquardt (L-M) algorithms to obtain thetransforming matrices. The SIFT is used to detect plenty ofdistinctive invariant features which can be used for reliablematching between the images with overlapping regions. Afterobtaining a series of pairwise matching points, we use the L-Malgorithm, a non-linear optimization, to obtain the translationand rotation parameters between two adjacent images. Using theparameters, any pair of two adjacent images can be transformedinto one coordinate system to generate a panorama. Preliminaryexperiments have been performed and the effectiveness of ourapproach is demonstrated.

作者: Liya Kong Qinghua Huang Shuohe Zheng Lianwen Jin Minhua Lu Siping Chen

作者单位: School of Electronics and Information engineering South China University of Technology Guangzhou, Ch Department of Biomedical Engineering, Medical School Shenzhen University Shenzhen, China

会议类型: 国际会议

会议名称: The 4th International Conference on Bioinformatics and Biomedical Engineering(第四届IEEE生物信息与生物医学工程国际会议 iCBBE 2010)

会议地点: 成都

会议语种:英文

页码: 1-4

在线出版日期: 2010-06-18（万方平台首次上网日期，不代表论文的发表时间）