Gaussian of Differences: A Simple and Efficient General Image Fusion Method

Abstract

The separate analysis of images obtained from a single source using different camera settings or spectral bands, whether from one or more than one sensor, is quite difficult. To solve this problem, a single image containing all of the distinctive pieces of information in each source image is generally created by combining the images, a process called image fusion. In this paper, a simple and efficient, pixel-based image fusion method is proposed that relies on weighting the edge information associated with each pixel of all of the source images proportional to the distance from their neighbors by employing a Gaussian filter. The proposed method, Gaussian of differences (GD), was evaluated using multi-modal medical images, multi-sensor visible and infrared images, multi-focus images, and multi-exposure images, and was compared to existing state-of-the-art fusion methods by utilizing objective fusion quality metrics. The parameters of the GD method are further enhanced by employing the pattern search (PS) algorithm, resulting in an adaptive optimization strategy. Extensive experiments illustrated that the proposed GD fusion method ranked better on average than others in terms of objective quality metrics and CPU time consumption.

SharDif: Sharing and Differential Learning for Image Fusion

“…Li et al [27] performed norm optimization on the fused images of MDLatLRR to obtain more significant fused images. Gaussian difference is used for image fusion, which is simple, efficient, and versatile [28].…”

“…Methods based on deep neural network feature representation learning include DenseFuse [40], RFNNest [43], FlFuse [51], PIAFusion [37], and PSFusion [44]. Model-based methods include GDFusion [28], and DDFM [52]. It should be noted that (1) All methods participating in the comparison use the weights and parameters given in the original papers to achieve the fusion of test data.…”

“…These metrics we selected include metrics that reflect the visual consistency of images, such as mutual information (MI) [53], Q AB/F (petrovic metric) [54], Q E (piella metric) [55], and feature mutual information (FMI) [56]. Indicators reflecting information fidelity, such as IFC [57], and other comprehensive indicators, such as structural similarity index measure (SSIM) [58], peak signal-to-noise ratio (PSNR), Q0 [58], and entropy (EN) [28], are included.…”

“…Qualitative comparison of all methods on six typical image pairs from TNO. From top to bottom: fusion results of DenseFuse [40], RFNNest [43], DDcGan [30], FLFuse [51], PIAFusion [37], PSFusion [44], GDFusion [28], DDFM [52], and our ShareDif.…”

Infrared and Visual Image Fusion Based on a Local-Extrema-Driven Image Filter

“…Subsequently, we selected eleven state-of-the-art image-fusion methods for comparison. These methods include the guided-filter-based image method (GFF) [36], the hybrid multiscale-decomposition-based image-fusion method (HMSD) [18], the Laplacian pyramid-and sparse-representation-based image-fusion method (LPSR) [25], the Gaussian of differencesbased image-fusion method (GDPSQCV) [37], the relative total variation-decompositionbased image-fusion method (RTVD) [38], the parameter-adaptive unit-linking dual-channel PCNN-based image-fusion method (PAULDCPCNN) [39], the GAN-based image-fusion method (FusionGAN) [16], the unified deep learning-based image-fusion method (U2Fusion) [40], the semantic-aware image-fusion method (SeAFusion) [28], and the representation learning-guided image-fusion method (LRR) [29]. For simplicity, we refer to our proposed local-extrema-driven filter-based image-fusion method as LEDIF.…”