The training of Karhunen–Loève transform matrix and its application for H.264 intra coding
- 89 Downloads
In H.264/AVC, 4 × 4 discrete cosine transform (DCT) is performed on the residual signals after intra prediction for decorrelation. Actually, residual blocks with different prediction modes exhibit different frequency characteristics. Therefore, the fixed transform matrix cannot match the energetic distribution of residual signals very well, which degrades the decorrelation performance. Fortunately, the energetic distributions of residual blocks with the same mode are relatively coincident, which makes it possible to train a universally good Karhunen–Loève transform (KLT) matrix for each mode. In this paper, an optimal frequency matching (OFM) algorithm is proposed to train KLT matrices for residual blocks and nine KLT matrices corresponding to nine prediction modes of 4 × 4 intra blocks are trained. Experimental results show that KLT with trained matrices yields a persistent gain over H.264 using 4 × 4 DCT with an average peak signal-to-noise ratio (PSNR) enhancement of 0.22dB and a maximum enhancement of 0.33dB.
KeywordsKarhunen–Loève transform Discrete cosine transform Intra coding H.264/AVC
- 1.Adaptive quantization encoding technique using an equal expected-value rule, JVT-N011.doc (2005) http://ftp3.itu.ch/av-arch/jvt-site/2005_01_HongKong, Hong Kong.
- 7.Halbach T (2002) Performance comparison: H.26L intra coding versus JPEG2000. JVT-D039Google Scholar
- 10.Lee YL, Han KH, Sim DG, Seo J (2006) Adaptive scanning for H.264 intra coding. ETRI J 28(5):668–671Google Scholar
- 13.Páta P (2006) Suboptimal Karhunen–Loève transform for compression of astronomical images. In: Proc. IEEE International Symposium on Signal Processing and Information Technology, Hawaii, USA, pp 332–337Google Scholar
- 17.Suh KB, Park SM, Cho HJ (2005) An efficient hardware architecture of intra prediction and TQ/IQIT module for H.264 encoder. ETRI J 27(5):511–524Google Scholar
- 18.Taubman DS, Marcellin MW (2002) JPEG2000: image compression fundamentals, standards and practice. Kluwer, BostonGoogle Scholar