A FUNDUS IMAGE MOSAICING STRATEGY BASED ON SPACE RESOLUTION INVARIANT DISCRETE SEQUENCE WAVELET TRANSFORM
基于恒定空间分辨率离散序列小波变换的眼底血管造影图像拼接方法

Space resolution invariant discrete sequence wavelet transform was applied to image mosaicing in fundus indocyanine green angiography (1CGA) thus removing the mosaicing seam without the decrease of space resolution caused by classic discrete sequence wavelet transform (DSWT) based 2. The average-weighted strategy was used to fit the wavelet low frequency parts togerther, and the direct average strategy to the high frequency parts to compromise visual effect and fidelity. In addition, a method of the precessing in wavelet field was used to avoid the influence on the fundus image resulted from inhomogeneous background light. The result of persent experiments is satisfactory.