Reflectance Transformation Imaging (RTI) is a photographic technique that enables the creation of a relightable digital representation of a physical object, providing enhanced surface details. It finds its early adoption in the cultural heritage fields, then expanding its application to the engineering domain to inspect manufactured surfaces. This literature review investigates the current state-of-the-art in RTI capture systems to support the development of an RTI microscope.
The review was carried out by researching papers in the IEEE and ScienceDirect databases, finding relevant connected papers using Researchrabbit, extracting key insights through a literature matrix, which was then converted into the literature review. The review addressed three core components: RTI setup, illumination strategy, and computational models. The two main RTI setups, Dome-based and Free-form, were compared in terms of costs, portability, automation, and accuracy.
The sample’s illumination was analysed under the quantity of light sources and their angles with respect to the sample. Lastly, the review evaluated three computational models: Polynomial Texture Mapping (PTM), Hemispherical Harmonics (HSH), and Discrete Modal Decomposition (DMD), highlighting the trade-offs between implementation complexity, computational cost, and surface reconstruction fidelity.
The results showed that RTI system components vary depending on the use-case priorities. The advantages of the Free-form setup are portability and accessibility, whereas the Dome-based setup is more accurate and automated. Regardless of the setup, the sample’s light should lie between 40 and 100 sources, with the lower angles (10-65 degrees) producing the most information. Lastly, regarding the model selection, DMD has the highest fidelity and computational costs, while PTM provides the lowest fidelity but is the simplest and least resource-intensive to implement.