PROJECT TYPE:

• Institutional Research Project

PROJECT DURATION:

• October 1, 2025 – September 30, 2029

PROJECT LEADER AND TEAM MEMBERS:

•  Assoc. Prof. Dr. Nicoletta Saulig – Project Leader
•  Asst. Prof. Dr. Siniša Miličić – Team Member
•  Assoc. Prof. Dr. Željka Tomasović – Team Member
•  Prof. Dr. Srđan Stanković – Team Member
•  Mr. Sc. Đorđe Stanković – Team Member

SHORT PROJECT DESCRIPTION:

The project “Photo-Analytical Entropy for Restoration and Emergence of Hidden Features and Structures” focuses on the development of advanced entropy–intensity methods for the quantitative analysis of digital images. Its aim is to enable accurate differentiation between noise and meaningful structural details, as well as to detect hidden features such as micro-cracks, degradation patterns, and intentionally inserted artifacts. This approach opens new possibilities in the digital restoration of artworks, where the identification of original layers and production phases is crucial, and in forensic material analysis, where it is necessary to reveal anomalies that are not visible to the naked eye.

The project includes a systematic theoretical elaboration and mathematical formalization of new measures, followed by their implementation in the form of efficient and scalable algorithms and testing on reference datasets. The computational complexity of the developed solutions is analyzed and compared with existing methods to ensure competitiveness and applicability in real-world conditions. Furthermore, collaborations with institutions in the fields of digital humanities, restoration, and industry enable the collection of authentic samples and the validation of results on diverse material contexts.

The final outcome of the project is a functional software prototype, accompanied by technical documentation and a dissemination plan for scientific and professional results. In this way, the project ensures the operational applicability of the developed methods in practice, contributing both to the advancement of image analysis techniques and to the preservation, interpretation, and forensic investigation of cultural and material heritage.

PROJECT OBJECTIVES

The objectives of the project focus on establishing both theoretical and practical foundations for advanced quantitative analysis of visual structures in digital images. Particular emphasis is placed on formulating entropy–intensity measures capable of describing the complexity, distribution, and organization of informational patterns. This involves developing mathematical models, examining their properties, and determining parameters that ensure stability, reliability, and sensitivity to subtle variations in image texture and structure.

In parallel with the theoretical development, the project aims to create computationally efficient algorithms that operationalize the defined measures. These algorithms will be optimized to handle images of varying resolution and scale, with controlled resource usage and the ability to integrate into existing digital workflows. Validation of the developed solutions will be carried out on diverse datasets, including synthetic samples and real digitized artifacts, in collaboration with domain experts from culture, digital humanities, and material analysis.

Alongside methodological advancement, the project prioritizes the visibility and long-term applicability of its outcomes. This includes the development of a functional software prototype, the preparation of technical and user documentation, dissemination of results through scientific and professional channels, and knowledge transfer activities directed toward cultural institutions, the research community, and the wider public.

PROJECT HOLDER: Juraj Dobrila University of Pula