Content-based medical image retrieval using compressed domain properties in the HEVC standard

Document Type : Original Article

Authors

1 Department of Computer Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran

2 Department of computer engineering, Mobarakeh branch, Islamic Azad University, Isfahan, Iran

Abstract

: In today's era, medical images play a very important role in providing better diagnostic services and medical research, and there is a need to retrieve required images from a sea of database images. In this article, a new method using the HEVC compression standard for compression and retrieval of medical images is presented in order to reduce storage space and increase efficiency. The important challenge in this field is the time-consuming nature of image retrieval.. In the proposed method, the database images are losslessly and intra-frame compressed with HEVC standard and the size of prediction blocks are extracted from it. Then it is normalized and their block graph correlation histogram is saved as a feature vector. The described operation is also performed on the query image, and its block correlation histogram is compared with the stored features, and as search results, the most similar images with the query image are displayed to the user. All existing methods in image retrieval in the compressed field with the HEVC standard have used the histogram of features, while in this article a new concept in the name of correloblock is presented to improve the accuracy of retrieval compared to other methods. The results of the experiments show that the proposed method has a higher performance accuracy than other similar methods and can be used to recover medical images.

Keywords

Main Subjects

Advanced Signal Processing

Articles in Press, Accepted Manuscript
Available Online from 23 November 2022

Files

Share

How to cite

Statistics