Identification And Analysis Of Tibetan Mani Stone Carving Based On Image Processing And Machine Learning

Mani stone carving is an important part of Tibetan culture and religion.It is an important basis for studying the development and religious distribution of Tibetan people.The Tibetan people carved the six-character mantra,the eight-character mantra and the Buddha image in Buddhism to form Mani.The stone pile landscape,the large stone pile is even composed of hundreds of millions of Mani stones.Mani stone carvings not only have a wide variety of engraving content,but also have various carving methods,which have certain historical value and artistic value.The related research of Mani stone carving is of great significance for understanding the development of Tibetan humanities.Due to the large number,it is difficult to achieve classification and statistics by manual means.Through image processing and machine learning,the automatic identification and classification of Mani stone inscriptions can be realized,which is of great significance for the digital protection and research of Mani stone inscriptions.In this paper,the depth image processing method and deep neural network algorithm are deeply studied,and a more complete Mani stone image data set is established.The main contents are as follows:1.Using the depth image to segment and extract the position of the Mani stone in the photo,with the depth camera capturing the collected Mani stone samples,and use the edge detection operator to complete the segmentation.Mark the location of the Mani stone inscriptions to create a tagged Mani stone training data set.2.Establish a neural network to train image data containing Mani stone inscriptions.The network structure of this paper contains 24 convolutional layers and 2 fully connected layers.The convolutional layer is used to extract the features in the image,and the next 2 full connections.The layer is used to judge whether the objects existing in the picture contain Mani stones,and are classified according to the corresponding labels.The classification here is mainly the classification of the engraving contents and techniques of the Mani stone inscriptions.After the test,the collected Mani stone image data were identified and analyzed.3.Using the classification result of the neural network model to establish the image database of Mani stone carving,which stores the image files,label files and related attributes of Mani stone inscriptions.This can be applied as the training data set and test data set of Mani stone carving in the study of Mani stone inscriptions.