Block-Based Multi-Scale Image Rescaling

TL;DR

This paper introduces a block-based multi-scale framework for image rescaling that improves super-resolution quality on high-resolution images by dynamically allocating scaling rates to image sub-blocks.

Contribution

The paper proposes a novel BBMR framework with dynamic sub-block scaling and a joint super-resolution method to better handle high-resolution images.

Findings

Significantly improves super-resolution quality on 2K and 4K images.

Effectively eliminates blocking artifacts in rescaled images.

Outperforms existing image rescaling methods in experiments.

Abstract

Image rescaling (IR) seeks to determine the optimal low-resolution (LR) representation of a high-resolution (HR) image to reconstruct a high-quality super-resolution (SR) image. Typically, HR images with resolutions exceeding 2K possess rich information that is unevenly distributed across the image. Traditional image rescaling methods often fall short because they focus solely on the overall scaling rate, ignoring the varying amounts of information in different parts of the image. To address this limitation, we propose a Block-Based Multi-Scale Image Rescaling Framework (BBMR), tailored for IR tasks involving HR images of 2K resolution and higher. BBMR consists of two main components: the Downscaling Module and the Upscaling Module. In the Downscaling Module, the HR image is segmented into sub-blocks of equal size, with each sub-block receiving a dynamically allocated scaling rate while maintaining a constant overall scaling rate. For the Upscaling Module, we introduce the Joint Super-Resolution method (JointSR), which performs SR on these sub-blocks with varying scaling rates and effectively eliminates blocking artifacts. Experimental results demonstrate that BBMR significantly enhances the SR image quality on the of 2K and 4K test dataset compared to initial network image rescaling methods.