Title :
Image registration based on CEMD
Author :
Ho-Hsuan Chang ; Jin-Shing Shia ; Zeng-Hwa Tsai
Author_Institution :
Dept. of Commun. Eng., I-Shou Univ., Kaohsiung, Taiwan
Abstract :
Image registration is a mean of image alignment among different images, thus it is a critical step in determining the performance of image data fusion or reconstruction applications such as image super-resolution. Image registration is still a challenging research topic and draws much attention these days. In this paper a new image registration technique based on CEMD is introduced and analyzed. The CEMD algorithm is the extension of EMD which is originated from the development of HHT that consists of EMD and HAS two parts. The HHT is characterized by the ability of analyzing nonlinear and non-stationary signals and is widely applied in many fields, such as: earthquake engineering, system identification, structural damage detection and health monitoring.
Keywords :
Hilbert transforms; image fusion; image reconstruction; image registration; image resolution; CEMD algorithm; HHT; Hilbert-Huang transform; complex empirical mode decomposition; earthquake engineering; health monitoring; image alignment; image data fusion; image reconstruction; image registration technique; image superresolution; nonlinear signal analysis; nonstationary signal analysis; structural damage detection; system identification; Films; Image registration; Image resolution; Next generation networking; Noise; Signal resolution; Simulation;
Conference_Titel :
Next-Generation Electronics (ISNE), 2013 IEEE International Symposium on
Conference_Location :
Kaohsiung
Print_ISBN :
978-1-4673-3036-7
DOI :
10.1109/ISNE.2013.6512418