CAJ | 학술논문

综合利用、处理多源遥感图像数据的融合理论与方法的需求越来越大,其中配准是遥感图像融合中重要的预处理步骤.分析了图像配准三类方法的特点和存在的问题,研究了遥感图像融合与图像配准位置之间的关系,发现了融合后图像的平均梯度随配准偏差变化的规律.在此基础上,将图像配准与后续的图像融合结合起来考虑,提出了基于融合结果进行准确配准的新方法.为了降低计算复杂性,定义了新的融合图像和平均梯度的计算公式.实验结果验证了该新方法的快速有效性,适用于同质遥感图像的配准.
종합이용、처리다원요감도상수거적융합이론여방법적수구월래월대,기중배준시요감도상융합중중요적예처리보취.분석료도상배준삼류방법적특점화존재적문제,연구료요감도상융합여도상배준위치지간적관계,발현료융합후도상적평균제도수배준편차변화적규률.재차기출상,장도상배준여후속적도상융합결합기래고필,제출료기우융합결과진행준학배준적신방법.위료강저계산복잡성,정의료신적융합도상화평균제도적계산공식.실험결과험증료해신방법적쾌속유효성,괄용우동질요감도상적배준.
Demands for data fusion theory and method has been increased much in synthetically using and processing the multi-remote-sense images,and registration is the important pretreatment step of images fusion.This paper analyses the characteristics and the problems about the three kinds of existing registration approaches,researches the connection between the image fusion and registration,discovers thedisciplinarian that the average grade of the fusion image changed with the registration error,and presents the idea that images registration using the fusion result, and then defines new fusion and average grade formulations to reduce the calculation cost.The experiment shows that the new registration method is fast and effective.and can be applied to the homogeneity remote-sense images registration.