Heritage Identification of Monuments using Deep Learning Techniques

Abstract

India is a nation with a plethora of cultural landmarks, including notable architectural masterpieces, 37 of which are UNESCO World Heritage master pieces. We must protect cultural heritages because they bind successive generations together over time. Architectural Designers, researchers, and travellers, etc. visit numerous historical locations, where it is frequently challenging for them to recognise and learn more about the historical significance of the monument in which they are interested. Due to the size and dependability of the information, the work of archiving, recording, and sharing the knowledge of these cultural assets is difficult. Modern machine learning and deep learning algorithms, as well as high processing computational resources, offer practical answers. The classification ofMonument satellite images or photographs can be atomized by utilising the Convoluted Neural Network techniques. During our project implementation, monuments images dataset was created with the help of google earth images. We have applied various pre-processing techniques and from pre-processed images we extracted features using feature extraction techniques such as Local Binary Patterns(LBP), Mean Standard Deviation(MSD). A model was developed using deep learning algorithms such as the Convoluted neural network(CNN). The results of our project are discussed in this paper. Initially we upload the monument satellite image, then the system recognizes the monument first then predicts whether the monument is heritage or not and gives the accuracy value as well as displays the monument image along with information of the monument.