iBlurDetect: Image Blur Detection Techniques Assessment and Evaluation Study

Roxanne A. Pagaduan, Ma. Christina R. Aragon, Ruji P. Medina

Abstract

The quality of images is essential in computer vision, image processing, and other related fields. Image restoration is one of the categories in image processing, where the quality of an image plays a vital role in the process. Blur detection is a pre-processing stage in image restoration. Using different blur detection techniques, the quality of an image can identify if blurry or not. This study aims to provide a comparative performance of the available state-of-the-art blur measure operators or blur detection techniques. Python 6.3 was used for testing and evaluating the blur detection techniques. Providing the confusion matrix, precision, recall, f-measure, accuracy, and execution time were used to compare blur detection techniques. In testing, the Gaussian kernel and threshold value were set to measure the performance of each technique. Provided on the evaluation results, in terms of accuracy rate, HWT leads the best result. Based on the computed scores, FFT got the highest precision score, while LAP got the highest recall score, and HWT got the highest f-measure score. In terms of the execution time, MLAP performs the fastest processing time among them all. Likewise, results of this study can use as resources before performing the image restoration.

Download


Paper Citation


in Harvard Style

Pagaduan R., R. Aragon M. and Medina R. (2020). iBlurDetect: Image Blur Detection Techniques Assessment and Evaluation Study.In Proceedings of the International Conference on Culture Heritage, Education, Sustainable Tourism, and Innovation Technologies - Volume 1: CESIT, ISBN 978-989-758-501-2, pages 286-291. DOI: 10.5220/0010307702860291


in Bibtex Style

@conference{cesit20,
author={Roxanne A. Pagaduan and Ma. Christina R. Aragon and Ruji P. Medina},
title={iBlurDetect: Image Blur Detection Techniques Assessment and Evaluation Study},
booktitle={Proceedings of the International Conference on Culture Heritage, Education, Sustainable Tourism, and Innovation Technologies - Volume 1: CESIT,},
year={2020},
pages={286-291},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010307702860291},
isbn={978-989-758-501-2},
}


in EndNote Style

TY - CONF

JO - Proceedings of the International Conference on Culture Heritage, Education, Sustainable Tourism, and Innovation Technologies - Volume 1: CESIT,
TI - iBlurDetect: Image Blur Detection Techniques Assessment and Evaluation Study
SN - 978-989-758-501-2
AU - Pagaduan R.
AU - R. Aragon M.
AU - Medina R.
PY - 2020
SP - 286
EP - 291
DO - 10.5220/0010307702860291