The direct use of curvelets in multifocus fusion

Abstract

In this effort, a data-driven and application independent technique to combine focal information from different focal planes is presented. Input images, acquired by imaging systems with limited depth of field, are decomposed using a relatively new analysis tool called curvelets. The extracted curvelets are representative of polar 'wedges' from the frequency domain. Fusion is performed on medial and peripheral curvelets by relevant fusion rules and the fused image combines information from different focal planes, while extending the depth of field of the scene. The main contribution in this effort is the direct use of curvelets in combining multifocal images. Several illustrative examples and objective comparisons are provided.Index Terms-Multifocus fusion, image fusion, curvelet transform, extending depth of field

Fusing Infrared and Visible Images via a First-Order Model

“…Over the past few decades, several techniques have been proposed for pixel-level fusion. These include the Laplacian pyramid (LP) [5,41,43], the discrete wavelet transform (DWT) [11], the dual-tree complex wavelet transform (DTCWT) [20,21], the curvelet transform (CVT) [16,33], the non-subsampled contourlet transform (NSCT) theory [13,14], the multi-resolution singular value decomposition (MSVD) [32], guided filtering fusion (GFF) [24], autoencoder-based approaches [15], and other techniques [38]. Recently, deep learning-based fusion methods have also been developed, including the DenseFuse method [22], the RFN-Nest method [23], the SDNet method [49], the SeAFusion method [40], image fusion based on proportional maintenance of gradient and intensity (PMGI) [50], image fusion based on convolutional neural network (IFCNN) [17,51], and fusion method based on generative adversarial networks (FusionGAN) [26].…”