Cutting, Shifting and Pasting for manipulation in DCT Domain

An approach to get the block DCT coefficients from one size to any size at arbitrary position of an image/video frame in the DCT domain directly by simple cutting, shifting and pasting operations is proposed. Many existing approaches, which require the input block(s) to align with the border of the output block, could be considered as special cases of the proposed approach. The practicality of the approach is illustrated by an example where an image can be resized arbitrarily in the DCT domain directly.