A New Approach for Effective Medical Image Retrieval

Authors

  • S. Vyshali
  • M. V. Subramanyam
  • K. Soundara Rajan

Abstract

The dominant rotated local binary pattern and histogram of oriented gradients are combined together which is useful to search the features of a medical image from a huge database. Similarity search has been done using hierarchical extreme learning machine of multi-layer perceptron and is merged with evolutionary algorithm to retrieve a medical image. This searching process is very easy, fast responsive nature and efficient for a new doctor or a doctor in a remote place to diagnose the patient. The dominant direction and histogram oriented gradients help the retrieving process to come with remarkable results. The quality of the retrieving process is tested for database and compared with state of medical image retrieval models. Content-based image retrieval is the application of computer vision techniques to the problem of digital image search in large databases. In this work, an image retrieval system based on curvature coding is focused.

Keywords: Quality parameters, curvature, retrieval, texture.

Downloads

Published

2020-01-08

Issue

Section

Articles