# Image rectangling network based on reparameterized transformer and assisted learning

## TL;DR

This paper introduces a new method for making stitched images have regular rectangular shapes without distorting their content.

## Contribution

A novel image rectangling approach using a reparameterized transformer and assisted learning with efficient local deformation.

## Key findings

- The proposed method achieves state-of-the-art performance in image rectangling.
- It maintains high content fidelity with a low number of parameters.
- A local thin-plate spline Transform strategy improves parallel efficiency.

## Abstract

Stitched images can offer a broader field of view, but their boundaries can be irregular and unpleasant. To address this issue, current methods for rectangling images start by distorting local grids multiple times to obtain rectangular images with regular boundaries. However, these methods can result in content distortion and missing boundary information. We have developed an image rectangling solution using the reparameterized transformer structure, focusing on single distortion. Additionally, we have designed an assisted learning network to aid in the process of the image rectangling network. To improve the network’s parallel efficiency, we have introduced a local thin-plate spline Transform strategy to achieve efficient local deformation. Ultimately, the proposed method achieves state-of-the-art performance in stitched image rectangling with a low number of parameters while maintaining high content fidelity. The code is available at https://github.com/MelodYanglc/TransRectangling.