Video encoding device, video encoding method, and program
AssigneeNEC Corporation
InventorKeiichi Chono
Life of this patent
Estimating the term…
Abstract
A video encoding device uses fractional pixel accuracy motion compensation prediction based on a DCT interpolation filter, and includes: a fractional pixel accuracy motion vector search unit 10 for estimating a fractional pixel accuracy motion vector using an interpolation filter 11 different from a DCT interpolation filter 21; and a fractional pixel accuracy motion compensation prediction signal updating unit 20 for updating a prediction signal of a position of the fractional pixel accuracy motion vector estimated by the fractional pixel accuracy motion vector search unit 10, with a prediction signal based on the DCT interpolation filter 21.
Claims3
A video encoding device that uses fractional pixel accuracy motion compensation prediction based on a DCT interpolation filter, the video encoding device comprising:
- a memory storing instructions;
- and
- a processor configured to execute the instructions to:
- estimate a fractional pixel accuracy motion vector using an interpolation filter different from the DCT interpolation filter;
- and
- update a prediction signal of a position of the estimated fractional pixel accuracy motion vector, by a prediction signal based on the DCT interpolation filter;
- wherein coefficients of the interpolation filter different from the DCT interpolation filter are [1, −4, 27, 9, −3, 1]/32 or [−2, −27, 8, −1]/32 at a fractional pixel position of 1/4, [1, −5, 20, 20, −5, 1]/32 or [−2, 18, 18, −2]/32 at a fractional pixel position of 2/4, and [1, −3, 9, 27, −4, 1]/32 or [−1, 8, 27, −2]/32 at a fractional pixel position of 3/4.
A video encoding method that uses fractional pixel accuracy motion compensation prediction based on a DCT interpolation filter, the video encoding method comprising:
- estimating a fractional pixel accuracy motion vector using an interpolation filter different from the DCT interpolation filter;
- and
- updating a prediction signal of a position of the estimated fractional pixel accuracy motion vector, by a prediction signal based on the DCT interpolation filter;
- wherein coefficients of the interpolation filter different from the DCT interpolation filter are [1, −4, 27, 9, −3, 1]/32 or [−2, −27, 8, −1]/32 at a fractional pixel position of 1/4, [1, −5, 20, 20, −5, 1]/32 or [−2, 18, 18, −2]/32 at a fractional pixel position of 2/4, and [1, −3, 9, 27, −4, 1]/32 or [−1, 8, 27, −2]/32 at a fractional pixel position of 3/4.
A non-transitory computer readable information recording medium storing a video encoding program that uses fractional pixel accuracy motion compensation prediction based on a DCT interpolation filter, when executed by a processor, performs a method for:
- estimating a fractional pixel accuracy motion vector using an interpolation filter different from the DCT interpolation filter;
- and
- updating a prediction signal of a position of the estimated fractional pixel accuracy motion vector, by a prediction signal based on the DCT interpolation filter;
- wherein coefficients of the interpolation filter different from the DCT interpolation filter are [1, −4, 27, 9, −3, 1]/32 or [−2, −27, 8, −1]/32 at a fractional pixel position of 1/4, [1, −5, 20, 20, −5, 1]/32 or [−2, 18, 18, −2]/32 at a fractional pixel position of 2/4, and [1, −3, 9, 27, −4, 1]/32 or [−1, 8, 27, −2]/32 at a fractional pixel position of 3/4.
Dependencies are inferred from the claim text.