Tourism scene classification based on cascade classifier

This paper proposes a two-layer cascade classifier based on global structural features and local texture features for tourism scene classification. The first layer classifier uses the global structural features to classify all the scene images. Then, a parameter δ is generated to distinguish the similarity between categories. Finally, the second layer classifier uses the local combined texture features to classify the scene images in the categories with high similarity. Experiment results show that this cascade classifier has good performance in improving the tourism scene classification accuracy.