A Novel Automatic System for Logo Based Document Image Retrieval Using Hybrid SVDM-DLNN

Abstract

Many government and private organisations represent themselves to the public using their own symbols or logo which is unique from others so that anyone can easily identify their products or belongings. This gives an ownership and source documentation to the owner by simply providing such logos. Using these logos for document retrieval in World Wide Web is a booming research in present era. Since usage of virtual documentation is increased day by day and handling this large data becomes a problem while searching for single data. In present research arena various document image retrieval models is available based on classification and clustering techniques. In this graph techniques are used to identify the issues in the automatic logo detection model using back propagation neural network along with the single value decomposition model. This proposed research model concerned about the document retrieval system based on the logo matching process to attain better efficiency and accuracy than the earlier detection models.