论文部分内容阅读
基于特征图像配准算法广泛应用于遥感、制导、医学等领域,用于配准不同传感器获取图像。配准流程中特征的定位准确与否直接影响到配准算法的有效性。在配准过程中引入超分辨率技术对待配准图像进行重建,可以恢复图像高频信息,仿真结果表明超分辨率后特征点的定位更准确,与参考图像配准精度得到提升。对红外、毫米波、可见光等不同波段的图像之间精确配准意义重大。“,”Feature-based image registration has always been used in remote sensing, weapon guidance, medical image analysis and so on, to align the images from different sensors. Localization accuracy directly affects the performance of the registration algorithm. Using super-resolution algorithm to reconstruct the sensed image can regenerate high frequency information, the experiment results prove that super resolution processing can decrease localization error and promote the accuracy of registration with reference image. The idea is useful in high-accuracy image registration between images from different wave bands, such as infrared, millimeter and visible light.