Selection of compression test images using variance-based statistical method

Chems El Houda Allaoui (LTIT Laboratory, Tahri Mohammed University of Bechar)
Abdesselam Bassou (LTIT Laboratory, Tahri Mohammed University of Bechar)
Ismahane Benyahia (LTIT Laboratory, Tahri Mohammed University of Bechar)
Mustapha Khelifi (LTIT Laboratory, Tahri Mohammed University of Bechar)

Indonesian Journal of Electrical Engineering and Computer Science
Vol 16, No 1: October 2019

In this research paper, we used a variance-based statistical method to select 20 test images, according to these latter types (natural, satellite and medical) among a sample of 300 images (100 natural, 100 satellite and 100 medical). The images selection has been done using parameters of image quality, namely PSNR, MSSIM and VIF, which were applied on three compression algorithms (DWT+SPIHT, JPEG, and JPEG2000).