A Novel Logo-Based Document Retrieval Using Hybrid Fuzzy Based CSA

Logo based document analysis plays an important role in many organizations for collecting information from massive number of administrative documents so that it can be summarized easily. Many researches are going on for improving the excellence of system, but the issues increase as the logos are similar to each other with minor differences. Conventional methods would not suitable for such complex process of identifying exact match so optimized models plays a vital role in this process. Cuckoo search algorithm is used in many cluster-based applications and it provides better convergence results than other optimization models. This proposed research model uses hybrid cuckoo search algorithm using global search procedure for enhancing its performance in analysing the logo-based document retrieval from the data set and proves its effectiveness in terms of fitness function and classification accuracy.