| Yazarlar |
Dr. Öğr. Üyesi Serkan KESER
Kırşehir Ahi Evran Üniversitesi, Türkiye |
| Özet |
| Karhunen-Loeve Transform (KLT) is generally not preferred much because it is a signal-dependent transform in signal compression. However, by developing effective algorithms, image compression that provides good performance can be achieved. In this study, two effective image compression methods based on eigenvector matrices obtained using KLT are applied. Image compression is performed by creating small-sized codebooks for the two methods. The first method is based on the number of eigenvector matrices used in the training phase. Some of the highest numbers of used eigenvector matrices are used for image compression. The second compression method uses eigenvectors of autocorrelation matrices for quantization. In this approach, quantization is performed using the principal components of the eigenvector matrices. Various codebook sizes have been tested for image compression. The qualities of the reconstructed test images were compared with DCT-based JPEG and Wavelet Transform-based JPEG2000 compression methods using the PSNR metric. Experimental results show that the PSNR values of the proposed methods give better results than that of JPEG. |
| Anahtar Kelimeler |
| codebooks | eigenvector matrices | Image compression | KLT |
| Makale Türü | Özgün Makale |
| Makale Alt Türü | SCOPUS dergilerinde yayımlanan tam makale |
| Dergi Adı | El-Cezeri: Journal of Science and Engineering |
| Dergi ISSN | 2148-3736 |
| Dergi Tarandığı Indeksler | Scopus |
| Makale Dili | İngilizce |
| Basım Tarihi | 01-2022 |
| Cilt No | 9 |
| Sayı | 2 |
| Sayfalar | 424 / 435 |
| Doi Numarası | 10.31202/ecjse.951417 |
| Makale Linki | http://dx.doi.org/10.31202/ecjse.951417 |
| Atıf Sayıları |
