Research On Classification Rule Of Images Emotion Extraction Based On Fuzzy Decision Tree

With the development of society and the network technique, a good deal of images'information has been produced. The way of sorting, storing and searching images is always a hot spot to be researched in the field of computers now.Because classification of images emotion could improve the accuracy of the images classification, people attach more and more attention to it. The main idea of the way of classification based on images emotion is to find out the mapping between images'vision features and images'semantic information, and excavate the relation between images'vision features and high semantic level in order to classify and label the images collection by semantic information.The key finding out the mapping between images'vision features and images'semantic information is finding out the classification rule. The mapping between images'vision features and images'semantic information is complicated, vagueness, flexible and loose. However, fuzzy decision tree could solves all the vagueness and flexible problem, could make the mapping between images'vision features and images'semantic information is more precise, and could extract the rule between images'vision features and images'semantic information. For the reason above, it get a better favorbale effect to use fuzzy decision tree to extract classification rule of images emotion, and acquire a rule set.A lot of related records had been read, in order to study the fuzzy decision tree which was used to extract images'emotion classification rule. In the fuzzy decision tree, the processing of the images'attributes and the algorithm used in classification rule of images emotion extraction have been bring up, such as the methods of extract and blur the images'vision features, the methods of express and blur the images'emotion features, the amelioration of the fuzzy decision tree algorithm used in classification rule of images emotion extraction, and the methods of extract classification rule of images emotion and acquire a rule set. Finally, a kind of fuzzy decision tree algorithm was presented, which was used to extract images'emotion classification rule which had a higher precision.