Open AccessImage Quality Assessment of No Reference JPEG Compressed Images Using Various Spatial Domain Features
Author(s) -
Shubham Anjankar,
Ajinkya M. Pund,
Parag Jawarkar
Publication year - 2021
Publication title -
international journal of next-generation computing
Language(s) - English
Resource type - Journals
eISSN - 2229-4678
pISSN - 0976-5034
DOI - 10.47164/ijngc.v12i5.440
Subject(s) - jpeg , computer science , distortion (music) , image quality , artificial intelligence , computer vision , spatial analysis , enhanced data rates for gsm evolution , quality (philosophy) , pixel , data compression , image compression , image (mathematics) , data mining , image processing , mathematics , statistics , telecommunications , amplifier , philosophy , bandwidth (computing) , epistemology
In today's world of technological advancement higher data rate and data transmission with minimal memoryrequirement is gaining importance. At the same time it also important to preserve quality of data to be transferredfrom various types of distortions and to assess the quality at the receiving end. Hence perceptual image qualityassessment is becoming more popular. This paper introduces a new approach in domain of no reference imagequality assessment of jpeg compressed images using spatial features .Considering the fact that pixel distortion,blurring and edge information are important aspects as far as distortions are concern. A novel approach is presentedin this paper to address the distortion categories. The results are obtained in conjunction with the subjectiveimage quality assessment of train and test image categories. Also results are compared with full reference imagequality assessment technique and parameters of full reference quality assessment. To achieve better accuracyalgorithm is tested on LIVE Texas' image database and our own developed database. Results are found to be wellcorrelating to the various parameters considered for the comparison purpose.