A new block-based DCT framework has been developed recently in[1] in which the first transform may choose to follow a direction other than the vertical or horizontal one - the default direction in the conventional DCT. In this paper, we focus on two diagonal directions because they are visually more important than other directions in an image block (except the vertical and horizontal ones). Specifically, we re-formulate the framework of two diagonal DCTs and use them in combination with the conventional DCT. We will discuss issues such as the directional mode selection and the cross-check of directional modes. Some experimental results are provided to demonstrate the effectiveness of our proposed diagonal DCT's in image coding applications. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Fu, J., & Zeng, B. (2006). Diagonal discrete cosine transforms for image coding. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4261 LNCS, pp. 150–158). Springer Verlag. https://doi.org/10.1007/11922162_18
Mendeley helps you to discover research relevant for your work.