Image Classification Using No-balance Binary Tree Relevance Vector Machine

Nowadays, Image classification method has been widely researched in the world. In this paper, we prepare four building categories for database. Firstly we use the Gabor filter for image processing to extract the image features, and then divide the images to different subregions for histogram-based Gabor features. At last, for image classification, Support Vector Machine (SVM) and Relevance Vector Machine (RVM) are known to outperform classical supervised classification algorithms. SVM has excellent performance to solve binary classification problems. RVM could be more sparsity than SVM. A new method based on relevance vector machine— No-balance Binary Tree Relevance Vector Machine (NBBTRVM) is proposed to define a class in this classification task. NBBTRVM could do a good performance according to our experiment results.