Abstract—The CBIR is a process that retrieves images on the basis of shape, texture and color, etc that is basically features. It
works by retrieving images related to Query Image (QI) from big databases. In existing CBIR systems retrieval efficiency is
confined by extracting limited feature sets. Different
optimization algorithms like Particle swarm optimization, Genetic algorithm, etc have been used for optimization purpose which are very old algorithms. There are number of other algorithms that have been proposed after it that gives better optimized results. Afterwards there is need of classifier to retrieve exact set of the images that will increase the performance of the proposed system and meets the user’s
requirement. So, in this paper we have proposed a new approach
in which firstly image content is improved using Clahe then features are extracted using independent component analysis
and select the relevant features that reduces the semantic gap by learning discriminative features directly from the images.
Afterwards trained distinct medical image selected features
from dataset are optimized using Cuckoo search to improve the
accuracy and precision values. The proposed system performance is improved using Support vector machine classifier which meets the user expectation and increase the efficiency of the system. The proposed system is tested in MATLAB in terms of recall, Precision, etc various parameter. The DICOM dataset is used for testing purpose.
Index Terms—CBIR, DICOM, CUCKOO, CLACHE,
independent component analysis.
The authors are with R Inder Kumar Gujral Punjab Technical University, Kapurthala, Punjab, India (e-mail: palavibhangu@gmail.com).
[PDF]
Cite: Palwinder Kaur and Rajesh Kumar Singh, "An Efficient Approach for Content-Based Image Retrieval
Using Cuckoo Search Optimization," International Journal of Modeling and Optimization vol. 9, no. 2, pp. 77-81, 2019.
