Loading...

The effect of a two steps searching mechanism Using Feature Vectors Related to Image Class in Improving the Performance of CBIR System

Sherafati, Shima | 2016

2311 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49403 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Jamzad, Mansoor; Manzuri Shalmani, Mohammad Taghi
  7. Abstract:
  8. Nowadays, retrieval is an inseparable part of user activities and due to growing usage of Content-Based Image Retrieval (CBIR), it has become a hot and challenging research topic specially in the past decade. The most important challenge that retrieval systems (including CBIR systems) are facing is the semantic gap between abstractions in the user’s mind and what is searched. One of the ways of dealing with this challenge is getting more information from the user about what he needs and so decreasing the distance between user’s will and what he gives to search engine as the description of his need. In this research, the class of query image is supposed to be given. For using this information, we propose a preprocess for rescaling feature space based on query’s class and an optimization method for setting the values of its parameters. Then, we run a classic Content-Based Image Retrieval approach on data in rescaled space for finding similar pictures to query image. Evaluations show that our proposed method gives 146% increase in the F measure on Corel5K dataset. In addition, we propose a method for reducing the dimensionality of feature vector while considering the optimization of proposed preprocess. The experimental result show that with the same accuracy as the ordinary retrieval, our method can reach 95% sparsity in weight matrices and so we can reduce the dimensionality of feature space to 5% on average
  9. Keywords:
  10. Semantic Gap ; Dimensionality Reduction ; Feature Vector ; Content Based Retrieval ; Image Retrieval ; Images Classification ; Term Weighting

 Digital Object List

 Bookmark

...see more