The Ancient Tamil Script is one of the most well-known ancient writing systems in many parts of India with Epigraphic records found in Rock edicts and Stone Hero. Although these texts are a rich source of history, very few applications have been developed to detect and translate ancient Tamil script characters. This is due to the lack of proper data and very few experts in these ancient texts that were lost long ago. In addition, many stone texts are in a condition that requires image enhancement and sound removal after photography. The aim of this project is to improve the structure to deal with anonymous documents in a systematic way and to organize the data needed for future work. In addition, the aim of this project is to come up with a way that overcomes the challenges of dealing with stone texts to make it easier to translate, using Convolutional Neural Networks and other in-depth learning methods.