Image Processing and Image Mining using Decision Trees

Valuable information can be hidden in images, however, few research discuss data mining on them. In this paper, we propose a general framework based on the decision tree for mining and processing image data. Pixel-wised image features were extracted and transformed into a database-like table which allows various data mining algorithms to make explorations on it. Each tuple of the transformed table has a feature descriptor formed by a set of features in conjunction with the target label of a particular pixel. With the label feature, we can adopt the decision tree induction to realize relationships between attributes and the target label from image pixels, and to construct a model for pixel-wised image processing according to a given training image dataset. Both experimental and theoretical analyses were performed in this study. Their results show that the proposed model can be very efficient and effective for image processing and image mining. It is anticipated that by using the proposed model, various existing data mining and image processing methods could be worked on together in different ways. Our model can also be used to create new image processing methodologies, refine existing image processing methods, or act as a powerful image filter.