Abstract:One of the most challenging problems in reconstructing a high dynamic range (HDR) image from multiple low dynamic range (LDR) inputs is the ghosting artifacts caused by the object motion across different inputs. When the object motion is slight, most existing methods can well suppress the ghosting artifacts through aligning LDR inputs based on optical flow or detecting anomalies among them. However, they often fail to produce satisfactory results in practice, since the real object motion can be very large. In this study, we present a novel deep framework, termed NHDRRnet, which adopts an alternative direction and attempts to remove ghosting artifacts by exploiting the non-local correlation in inputs. In NHDRRnet, we first adopt an Unet architecture to fuse all inputs and map the fusion results into a low-dimensional deep feature space. Then, we feed the resultant features into a novel global non-local module which reconstructs each pixel by weighted averaging all the other pixels using the weights determined by their correspondences. By doing this, the proposed NHDRRnet is able to adaptively select the useful information (e.g., which are not corrupted by large motions or adverse lighting conditions) in the whole deep feature space to accurately reconstruct each pixel. In addition, we also incorporate a triple-pass residual module to capture more powerful local features, which proves to be effective in further boosting the performance. Extensive experiments on three benchmark datasets demonstrate the superiority of the proposed NDHRnet in terms of suppressing the ghosting artifacts in HDR reconstruction, especially when the objects have large motions.

Deep HDR Reconstruction of Dynamic Scenes

MergeNet: Single High Dynamic Range Image Reconstruction Method.

Deep high dynamic range imaging of dynamic scenes

Multi-Scale Dense Networks for Deep High Dynamic Range Imaging

Learning Regularized Multi-Scale Feature Flow for High Dynamic Range Imaging

Reconstructing HDR Image from a Single Filtered LDR Image Base on a Deep HDR Merger Network

Deep SR-HDR: Joint Learning of Super-Resolution and High Dynamic Range Imaging for Dynamic Scenes

Deep HDR Imaging via A Non-local Network

High Dynamic Range Imaging for Dynamic Scenes With Large-Scale Motions and Severe Saturation

Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenes

Reconstructing a High Dynamic Range Image with a Deeply Unsupervised Fusion Model

Multi-scale Contextual Attention Based HDR Reconstruction of Dynamic Scenes

A Unified HDR Imaging Method with Pixel and Patch Level

Deep progressive feature aggregation network for multi-frame high dynamic range imaging

HDR Video Reconstruction with a Large Dynamic Dataset in Raw and sRGB Domains

HDR-Net-Fusion: Real-time 3D dynamic scene reconstruction with a hierarchical deep reinforcement network

HDR Video Reconstruction: A Coarse-to-fine Network and A Real-world Benchmark Dataset

Deep Progressive Feature Aggregation Network for High Dynamic Range Imaging

Single Exposure High Dynamic Range Image Reconstruction Based on Deep Dual-Branch Network

Progressive and Selective Fusion Network for High Dynamic Range Imaging