PSO-WESRGAN: A NOVEL DOCUMENT IMAGE SUPER RESOLUTION

Abstract

One of the major challenges of document images that can hinder readability and the analysis of information is low resolution, typically caused by low-pixel density scanning or excessive compression to save storage space. This results in images lacking fine details, making it difficult to recognize important information. Super-resolution techniques are essential to addressing these issues. These techniques enhance image quality by increasing resolution while maintaining fine details. The PSO-WESRGAN is an innovative method, which combines wavelet processing, deep transfer learning, and particle swarm optimization (PSO). Wavelet processing analyzes image detail at diverse scales and orientations, while transfer-based deep learning advantages pre-trained models on vast image datasets. By integrating PSO, the method’s efficiency is enhanced through optimal exploration of the solution space to identify the best parameters for the super-resolution model. Experimental results demonstrate the effectiveness of this approach and pave the way for future advances in document image resolution.