Caption removal method for digital video editing system
AssigneeNational Central University
InventorsTsung-Han Tsai, Chih-Lun Fang
Life of this patent
Estimating the term…
Abstract
A caption removal method for a digital video editing system includes the following steps. A video frame of a video sequence is inputted. A high level executing stage is provided. The high level executing stage includes the step of recovering the structural region of the temporal domain in the video frame and the step of recovering the structural region of the spatial domain in the video frame. A low level executing stage is provided. The low level executing stage includes the step of recovering the non-structural region in the video frame.
Claims17
A caption removal method for a digital video editing system comprising the following steps:
- inputting a video frame of a video sequence;
- providing a high level executing stage, the high level executing stage comprising the following steps:
- recovering the structural region of the temporal domain in the video frame, further comprising:
- analyzing and describing the distribution of the structural region of the temporal domain to obtain a most similar block using a rotated block matching algorithm;
- obtaining an initial position of an interpolation block using an adaptive interpolation algorithm;
- and
- adjusting the coordinates of the interpolation block adaptively to obtain a desired recovery block using the rotated block matching algorithm;
- and
- recovering the structural region of the spatial domain in the video frame, further comprising:
- defining that the structural distribution of the spatial domain has characteristics similar to those of an SP-Line curve;
- performing the rotated block matching algorithm to obtain a first similar block and a second similar block with a fractional structural block;
- calculating the tangent vector at the center coordinate of the first similar block and the tangent vector at the center coordinate of the second similar block to obtain the SP-Line;
- and
- moving the first similar block to the fractional structural block along the SP-Line to generate the recovery structure;
- and
- providing a low level executing stage, the low level executing stage comprising the following step:
- recovering the non-structural region in the video frame, further comprising:
- determining whether the video frame is a first video frame;
- recovering the non-structural region in the video frame with smooth propagation when the video frame is determined as the first video frame;
- and
- recovering the non-structural region in the video frame with block duplication when the video frame is not determined as the first video frame.
Dependencies are inferred from the claim text.