Reduce Reference Image Quality Assessment for High Dynamic Range based on Tone-Mapping Operator
Total View This Week0
Total View This Week0
Institusion
Universitas Bakrie
Author
Syafa'ah, Salmaa Badriatu
Subject
Image Processing
Datestamp
2019-08-28 12:37:18
Abstract :
This thesis propose an objective image quality assessment for High Dynamic Range (HDR) images with reduced reference based on Tone Mapping Operator (TMO). As we know, HDR images can now be displayed on a standard device such as a smartphone. The valuation method here is classified as the Reduce-Reference (RR) method as only partial reference information is available. First, the TMO images and the reference images is extracted. Then, the images quality score is measured with the features extracted. This study use three main models i.e. Quality Assessment 1 (QA1), Quality Assessment 2 (QA2), and Quality Assessment 3 (QA3). To know how well this model, Pearson and Spearman correlation is applied. As the ground thruth data, public subjective score is utilized. The result show that feature combine is the best feature with QA1\_Cubed as the highest correlation.
This thesis propose an objective image quality assessment for High Dynamic Range (HDR) images with reduced reference based on Tone Mapping Operator (TMO). As we know, HDR images can now be displayed on a standard device such as a smartphone. The valuation method here is classified as the Reduce-Reference (RR) method as only partial reference information is available. First, the TMO images and the reference images is extracted. Then, the images quality score is measured with the features extracted. This study use three main models i.e. Quality Assessment 1 (QA1), Quality Assessment 2 (QA2), and Quality Assessment 3 (QA3). To know how well this model, Pearson and Spearman correlation is applied. As the ground thruth data, public subjective score is utilized. The result show that feature combine is the best feature with QA1\_Cubed as the highest correlation.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
