眼底循环障碍疾病以及全身性疾病导致的眼底改变均可不同程度地显征于眼底图像。在 ICG眼底造影结合共聚焦激光显微镜技术条件下 ,可得到高分辨率的眼底视网膜和脉络膜局部图像 ,而眼底病症可能散布于眼底各生理部位 ,眼底图像序列的自动拼接可解决分辨率和单次可观测视野间的矛盾。采用相位相关法匹配 ICGA眼底图像 ,应用基于遗传算法的非刚性光流场估计方法计算未建模的眼底微小运动 ,配准不完全匹配局部。文中方法完备了 ICGA眼底图像拼接方法 ,提高了精度和正确匹配率 ,实验中有效地克服图像虚影 ,保留了图像细节 ,结果良好 Fundus circulatory disorders and systemic diseases caused by fundus changes can be varying degrees of manifestations in fundus images. Under ICG fundus confocal laser microscopy, high resolution retinal and choroidal images can be obtained, while ocular conditions may be scattered in various physiological regions of the fundus. Automated stitching of the ocular fundus images can resolve the resolution and single Can be observed between the contradictions. The phase correlation method was used to match the fundus image of ICGA. The non-rigid optics flow field estimation method based on genetic algorithm was used to calculate the unmodelled fundus motion and the registration was incompletely matched. The method in this paper has completed ICGA fundus image stitching method to improve the accuracy and the correct matching rate, the experiment effectively overcome the image ghost, preserves the image details, the result is good