Blind Image Quality Assessment Score for Humanities Online Digital Repositories

The proliferation of digital image repositories has revolutionized the accessibility of visual data across various domains, from scientific research to cultural heritage preservation. However, the quality of images within these repositories varies significantly, posing a challenge for users relying on accurate and high-quality visual information. In recent years, a substantial effort has been made to find an objective and blind image quality assessment (IQA) metric that can correlate with perceived quality measurements. Unfortunately, there is still no definite method to asses this problem because of the complex nature of objectively measuring image distortions. In this paper, we present a comprehensive framework for assessing image quality in online repositories for humanities. The approach relies on a combination of known image processing techniques, including histogram analysis, artefact estimation and dynamic range evaluation. Other parameters are also taken into account, such as resolution, dots per inch (DPI), bit depth and format. By automatically grading image quality using these methods, we aim to provide researchers and practitioners within the humanities field, with a reliable tool for evaluating the trustworthiness and suitability of images sourced from national repositories. Through a systematic examination of image quality metrics, the strengths and weaknesses of existing image collections can be highlighted, paving the way for enhanced curation standards and improved access to high-quality visual resources. Methods were tested on images from online repositories in public domain.