Classification of High-resolution Images with Local Binary Pattern and Convolutional Neural Network: An Advanced Study

It is very important to accurately classify high-resolution satellite images and to classify each section of the image separately. Complex patterns, on the other hand, are difficult to identify. The deep learning method is used to deal with this challenge. The goal of the deep learning method is to extract a large number of features without the need for human intervention. Nonetheless, integrating deep features with texture characteristics improves classification performance. Deep feature learning mixed with texture-based classification is made easier with the suggested system. Local Binary Pattern (LBP) is used to extract textural features, whereas Convolutional Neural Network is used to extract deep features (CNN). The main objectives of the proposed system are: (1) To efficiently combine deep features with texture features. (2) To increase the classification accuracy. (3) To classify the land cover/land map area of the remote sensing image correctly. The suggested method is implemented, and the results are checked to ensure its efficacy. When texture features are incorporated with a deep learning approach, experimental results demonstrate that classification performance has improved.