Optimized Type - 3 Fuzzy Logic for Robust Learning Algorithm ‎in OCR Recognition

Optimized Type - 3 Fuzzy Logic for Robust Learning Algorithm ‎in OCR Recognition P. Sweetlin G. Jayalalitha

Optimized Type-3 Fuzzy logic system integrated with robust learning algorithm and enhanced with OCR performances. This approach is ‎specifically within the context of educational applications. OCR played an important role in converting printed educational materials into ‎digital formats. The reason for selecting fuzzy logic in the education sector is not only learning for the decision-making process. Fuzzy logic ‎is a more realistic and flexible evaluation. Results as more consistent, objectives. Type-1 fuzzy logic systems have single-valued ‎membership functions, they only handle basic-level operations. Type-2 fuzzy logic is computationally more complex with requires larger ‎resources to process in educational applications. To overcome this issue Optimized Type-3 Fuzzy logic combined the type-1 and type-2 ‎fuzzy logic systems with the terms to improve the adaptability and performance. Introducing the optimized Type-3 fuzzy logic algorithm ‎used PSO noted as PSO-OT3FL-RLA, helps to improve the accurateness and strength of the OCR system. This system also incorporates ‎uncertainty in the recognition process and ensures adaptable recognition outcomes. The proposed model is an advanced learning algorithm ‎integrated with fuzzy logic that effectively handles noisy input data. PSO technique ensures efficient exploration and maintains the balance ‎between the quality of the result and computational efficiency. Utilize the proposed approach performed through the traditional OCR ‎methods in terms of accuracy and computational error rate. After evaluating the overall performance of the proposed model, we provide a ‎result of 94.67%. It provides a promising solution for OCR-based applications in educational sectors‎.
07 08 2025 111 121 10.14419/608p4p44 https://www.sciencepubco.com/index.php/IJBAS/article/view/33602 https://www.sciencepubco.com/index.php/IJBAS/article/download/33602/18357 https://www.sciencepubco.com/index.php/IJBAS/article/download/33602/18357