| Yazarlar |
Dr. Öğr. Üyesi Serkan KESER
Kırşehir Ahi Evran Üniversitesi, Türkiye |
| Özet |
| This study proposes a new Karhunen-Loeve transform-based algorithm with acceptable computational complexity for lossy image compression. The proposed study includes a simple algorithm using downsampling and KLT. This algorithm is based on an autocorrelation matrix found by clustering the highly correlated image rows obtained by applying downsampling to an image. The KLT is applied to the blocks of the downsampled image using the eigenvector matrix of the autocorrelation matrix, and the transform coefficient matrix is obtained. One of the important features of the proposed method (PM) is sufficient for a test image to have one transform matrix with low dimensional. The PM was compared with JPEG, BPG, and JPEG2000 compression methods for the Peak signal-to-noise ratio- Human Visual System (PSNR-HVS) and the Structural Similarity Index Measure (SSIM) metrics. The mean PSNR-HVS values of the PM, JPEG, JPEG2000, and BPG were 37.44, 37.16, 37.45, and 39.14, respectively, and their mean SSIM values were 0.95, 0.93, 0.952, and 0.962, respectively. It has been observed that the PM generally gives better results than other methods for images containing low-frequency components at high compression ratios. As a result, PM can be used to compress images, especially those containing low-frequency components. |
| Anahtar Kelimeler |
| Makale Türü | Özgün Makale |
| Makale Alt Türü | Ulusal alan endekslerinde (TR Dizin, ULAKBİM) yayımlanan tam makale |
| Dergi Adı | Bitlis Eren üniversitesi fen bilimleri dergisi |
| Dergi ISSN | 2147-3129 |
| Dergi Tarandığı Indeksler | TR DİZİN |
| Makale Dili | İngilizce |
| Basım Tarihi | 03-2023 |
| Cilt No | 12 |
| Sayı | 1 |
| Sayfalar | 215 / 225 |
| Makale Linki | https://doi.org/10.17798/bitlisfen.1225312 |
