Multiclassification Hybrid Model Using CNN and SVM for National Emblems

The national emblem is the symbol of high respect for the nation, this symbol indicates the nation’s power, wealth, pride, and people. The national emblem has a special respect and place in both the nation’s history and even the nation’s present. These national emblems represent the identity of a nation. These national emblems are mainly used on occasions of national importance. So, this was important to preserve these emblems’ importance in this AI-growing modern era. So hybrid model is developed over 8 national emblems using CNN and SVM deep learning approaches. These 8 national emblems represent the eight countries, India, the UK, Germany, Australia, New Zealand, Italy, the USA, and the Netherlands. A dataset of 9950 images was used to train this hybrid model. 96.92% is the best accuracy obtained for this multiclassification of national emblems. 90.15% is the best F1 score recorded. And macro average is 86.0225% for this classification. In the future, more than 8 national emblems can be considered for multiclassification. And other hybrid models can be developed for obtaining better accuracy.