EpigraphiX-AI: Neural Epigraphical OCR, Topological Binarization, Strict Manuscript Authentication, and Multi-Model Intelligence for Degraded Historical Palm-Leaf Manuscripts

Historical South Indian palm-leaf manuscripts (Thaliyola), inscribed using incised iron styluses (Ezhanithandu), preserve invaluable ancient Sanskrit, Grantha, and Old Malayalam treatises spanning Ayurveda, astronomy, mathematics, and philosophy. However, severe physical degradation—including high-frequency cellulose fiber striations, biological decay, uneven stylus incision depths, and carbon ink dispersion loss—renders conventional OCR engines ineffective. In this paper, we present EpigraphiX-AI, an end-to-end epigraphical intelligence suite and neural optical character recognition architecture. The system introduces six core scientific innovations: 1. Strict Palm-Leaf Authentication: Multi-gamut HSV color space classifier and O(H+W) spatial span tracing that isolates authentic manuscripts while rejecting non-manuscripts (human portraits, outdoor scenes, digital UI) with 99.4% precision.2. Fiber-Aware Neural Inpainting (FANI 2.0): Directional Gabor filtering and Photometric Stereo normal estimation to isolate stylus groove incisions from fibrous wood grain textures.3. O(1) Integral Sauvola Binarization: Constant-time adaptive thresholding operating at sub-3.8ms latency per folio.4. Persistent Homology Betti Filtration: Simplicial complex topological filtering (Betti 0, Betti 1) preserving closed character loops in Grantha ligatures.5. 5-Model Decision Space: Comparative machine learning benchmark across SVM, Random Forest, Gaussian Naive Bayes, k-NN, and CNN Neural Lattices.6. Multilingual Semantic Bridge and Dynamic Canvas: Real-time interactive glyph editing and Grantha-to-Malayalam-to-English translation. Experimental evaluation on an archival corpus of 1,250 historical palm-leaf folios demonstrates that EpigraphiX-AI achieves a Word Accuracy Rate (WAR) of 97.4%, Character Accuracy of 98.6%, Character Error Rate (CER) of 1.4%, Precision of 98.6%, Recall of 98.4%, and Palm Authentication Accuracy of 99.4%.