Research On Fusion Algorithm Of UAV Infrared Image And Visible Image Registration

With the continuous development of multi-source imaging sensor technolo gy,single-source imaging equipment is difficult to meet the requirements of p ractical engineering.Applying multi-source sensors to UAV requires high quali ty images to ensure.Therefore,the registration and fusion technology of infrar ed images and visible light images is a hotspot in recent years.Using appropr iate registration and fusion techniques on UAV,high-efficiency and high-precis ion image data transmission can be achieved,and UAV can be improved in t he military.The role of the civilian sector.In this paper,the registration algor ithm and fusion algorithm of infrared image and visible image are studied.Th e main research contents are as follows:(1)Analyze the domestic and foreign registration and fusion technologies,and study the parameter optimization algorithm that maximizes mutual inform ation.According to the principle of Powell algorithm,the selection of initial p oint,the number of iterations and the influence of given error on the algorith m are analyzed experimentally.According to the improved random walk algori thm,the choice of step size and the number of iterations are analyzed experi mentally.influences.A hybrid optimization algorithm combining Powell and i mproved random walk algorithm is proposed.The performance of the optimiza tion algorithm is verified by experimental comparison.(2)Aiming at the shortcomings of low correlation between visible light i mage and infrared image,an image registration method based on edge region and cross-correlation is proposed to remove the low correlation between image s,and the high correlation is retained for registration.The similarity measure function and optimization algorithm greatly improve the effectiveness and succ ess rate of the registration.(3)Several traditional image fusion algorithms are studied: upgrade wavel et transform,principal component analysis PCA fusion algorithms,and Laplace pyramid fusion algorithnm.According to the characteristics of infrared image and visible light image taken by UAV imaging equipment,a fusion algorithm based on color space IHS transform and update wavelet transform is propose d.Several fusion algorithms are compared and analyzed.The results show that the algorithm can not only retain infrared The brightness information of the t arget can also retain the details of the visible image.