Incomplete Input Inferences on Fuzzy Expert Systems

Recently, the graphic interfaces are so friendly, and the vast multimedia information is growing so quickly. The studying of multimedia database is becoming a major class. Let the users can find the pictorial that they desired efficiently is the target of this research. There are two kinds of features embedded in the picture: the visual feature and the relationship feature. The visual feature includes color, texture and shape of objects. The relationship feature is the spatial relationships among the objects. In the studying of spatial relationship, previous related works all focus on the directional relationship, as 2D-String, 2D C-String and 2D-PIR, etc. In this research, we show that the distance is an important factor of spatial relationship. We propose a new representation to embed the distant relationship among objects. We also develop a new similarity measurement function and retrieval method based on our new representation. We use the R_norm measure to evaluate the quality of our similarity measurement method. We find that our new representation can work well in distant and directional relationship. The similarity retrieval method based on our new representation is more efficient and flexible for userpsilas request.