Research On Remote Sensing Image Fusion Base On The Second Curvelet Transform

Images and data in the same region can be obtained by different types of sensors. A satisfying image which contains rich useful information can be obtained by remote sensing image fusion techniques, which can combine the multi-images information. The image quality can be greatly improved by image fusion techniques. This will result in a good result to human visual perception or post-processing for computer. The second-generation curvelet transform is used for remote sensing image fusion in this paper. The curvelet transform can be used to efficiently extrude the features of original images and provide more information for fused image. Main works in this paper include:1) The performance of existing image fusion methods is compared and analyzed. The shortcoming of the existing image fusion methods is also pointed out.2) An efficient remote sensing image fusion method is proposed based on the second-generation curvelet transform. Source images are decomposed by the second-generation curvelet transform. Low frequency coefficients are fused by averaging method and high frequency coefficients are fused by local standard variance method. The proposed method is compared with some traditional image fused methods.3) In order to keep more spectrum information in the fused image, an efficient image fusion method which combines curvelet transform with HIS transform is proposed. Experimental results show that the proposed method can efficiently extrude the details in fused image while keep the spectrum information. The overall performance of the proposed method is better than that of the traditional image fused methods.4) Considering the visual characteristics of human eyes, an efficient remote sensing image fusion method by combing curvelet transform with the visual characteristics is proposed. The proposed method is compared with some traditional image fusion methods.