Research on Color Optimization in Intangible Cultural Heritage Digitalization Process Based on Deep Learning Algorithms

Le résumé fourni par la source

With society's transformation from industrialization and informatization to intelligence, intangible cultural heritage (ICH) is also experiencing a trend of “living heritage.” This research takes movable-type printing as an empirical study object, addressing the issues of distorted rendering effects and limited technical compatibility that affect immersion during the construction of interactive scenarios. To this end, we propose a Deep Learning Super Sample (DLSS) algorithm applicable to the digitalization process of ICH. This algorithm, based on in-depth analysis of scene rendering distortion and super-resolution reconstruction, employs Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) for rendering super-resolution and reconstruction. Combined with a color-attribute-based post-processing optimization method, the improved DLSS algorithm reconstructs a custom rendering shader, optimizing the clarity, contrast, and saturation of real-time rendering scenes, thereby enhancing the application value of ICH in the digital age. The experimental results show that the reconstructed DLSS algorithm significantly optimizes relevant data such as rendering frame rate and rendering time in real-time interactive scenes, thus verifying the feasibility and effectiveness of real-time rendering technology based on the deep learning super-sampling algorithm in the “living heritage” and innovative design of ICH.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.