Application of Deep Learning Strategy for Multi-classification of Indian Heritage Images

For the benefit of the world, heritage conservation and protection has been an active and important area of research. Heritage is considerably more than simply maintaining, unearthing, exhibiting, or restoring a collection of antiquated items. A heritage term covers a large range of inherited traditions, monuments buildings, stones, culture, birds, animals, and many more, which collectively depicts the quality of the legacy of a country. The preservation and digitization of cultural heritage are yet uncharted territories for many scholars therefore, an efficient heritage recognition and classification (HRC) model has been created based on deep learning (DL) strategy. The complete model conduction and implementation have been done using various Indian heritage images comprises of 10000 images containing four categories of heritage images namely animals, birds, monuments, and paintings. The complete process outcome resulted in a recognition accuracy of 94.32% using the binary classification technique, whereas 95.43% accuracy has been observed for further multi-classification tasks of heritage category images. The suggested study will assist in sustainable development, environmental upgradation, heritage preservation, and economic evolution for society and researchers in the aforementioned field.