Image Quality Prediction for DCT-based Compression

Ruslan Kozhemiakin , Vladimir Lukin , Benoit Vozel (1)

1 IETR - Institut d'Électronique et des Technologies du numéRique

a method for prediction and providing compressed image quality for lossy compression techniques based on discrete cosine transform (DCT) is proposed. A specific property of the designed method is its ability to predict compressed image quality with appropriately high accuracy using a limited number of analyzed blocks. This accelerates prediction of lossy compression quality substantially. The method is originally proposed for JPEG with uniform quantization and then generalized for other, advanced, DCT based coders AGU and ADCT.

Mots clés: lossy compression; JPEG; DCT; AGU; quality prediction and providing

Domaines: Réseaux et télécommunications [cs.NI]; Electronique

Ruslan Kozhemiakin, Vladimir Lukin, Benoit Vozel. Image Quality Prediction for DCT-based Compression. 14th International Conference on the Experience of Designing and Application of CAD Systems in Microelectronics (CADSM), Feb 2017, Svalyava, Ukraine. ⟨10.1109/CADSM.2017.7916121⟩. ⟨hal-01622073⟩