Method of temporal resampling and apparent motion speed change for motion picture data
The present invention is a method of processing a sequence of frames from digital motion picture cameras to allow remapping of the time-base of the input frames to a new and possibly variable time-base in the output frame sequence. This invention provides a method that makes small adjustments to the user's specified positions to achieve optimal integer mapping from input frames to output frames in the integrated function. The present invention provides for an algorithm for frame rate resampling providing a method for determining the mapping of input frames to output frames and a method for smoothly ramping the mapping of input frames to output frames based on interactive user input and a method for optimizing the alignment of input frames to output frames such that during areas of constant speed the alignment is optimally centered on the frames and a method for computing the weighting function for averaging input frames to output frames and a method for handling negative values in the weighting function in regions of bright highlights.
1. A method of processing digital motion picture data captured by a digital motion picture camera, the method comprising:
receiving the digital motion picture data from the digital motion picture camera, the received digital motion picture data including received digital input frame data;
receiving keyframe input data that at least partially defines specified digital input frame data and at least partially defines a desired output sequence,
wherein the specified digital input frame data is at least a portion of the received digital input frame data, wherein the specified digital input frame data is representative of a specified group of digital input frames, and wherein the specified group of digital input frames includes a specified starting digital input frame representative of a specified start position;
comparing the specified digital input frame data to the desired output sequence to detect whether the desired output sequence is a constant-speed output;
if the desired output sequence is detected to be constant-speed output, generating adjusted digital input frame data in order to adjust the specified start position to an adjusted start position and increase integer-to-integer mapping between the specified digital input frame data and the desired output sequence,
wherein the adjusted digital input frame data is representative of an adjusted group of digital input frames, the adjusted group of digital input frames including an adjusted starting digital input frame that differs from the specified starting digital input frame; and
generating digital output frame data using the adjusted digital input frame data, wherein the generating includes modifying the adjusted digital input frame data based at least in part on the desired output sequence.
2. The method of claim 1 , wherein the keyframe input data that at least partially defines additional specified digital input frame data and at least partially defines a second desired output sequence,
wherein the additional specified digital input frame data is at least a portion of the received digital input frame data, wherein the additional specified digital input frame data is representative of a second specified group of digital input frames, and wherein the second specified group of digital input frames includes a second specified starting digital input frame representative of a second specified start position.
3. The method of claim 2 , further comprising:
comparing the additional specified digital input frame data to the second desired output sequence to detect whether the second desired output sequence is a constant-speed output;
if the second desired output sequence is not detected to be constant-speed output, generating additional digital output frame data using the additional specified digital input frame data, wherein the generating includes modifying the additional specified digital input frame data based at least in part on the second desired output sequence.
4. The method of claim 3 , wherein the generating of the additional digital output frame data comprises computing a piecewise polynomial function based at least in part on the second desired output sequence.
5. The method of claim 4 , wherein the generating of the additional digital output frame data further comprises calculating weighted values based on weighted sums of portions of the additional specified digital input frame data representative of respective groups of the second specified group of digital input frames.
6. The method of claim 5 , wherein the generating of the additional digital output frame data further comprises adjusting the weighted values based on the piecewise polynomial function.
7. A method of processing digital motion picture data, the method comprising:
receiving the digital motion picture data generated by a digital motion picture camera, the received digital motion picture data being representative of a motion picture;
receiving a plurality of user-specified keyframes that define a plurality of specified regions of respective time periods of the motion picture;
detecting a rate-change region from among the plurality of specified regions, the rate-change region comprising a plurality of digital input frames to be time-based modified as a digital output frame; and
computing the digital output frame based at least in part on a piecewise function that changes depending on a brightness level of the plurality of digital input frames.
8. The method of claim 7 , wherein the piecewise function is a function of a sum of a plurality of weighted values.
9. The method of claim 8 , wherein the plurality of weighted values is based at least in part on respective ones of the plurality of digital input frames.
10. The method of claim 7 , further comprising detecting a steady-rate region from among the plurality of specified regions, the steady-rate region comprising a second plurality of digital input frames to be output as digital output frames without being time-based modified.
11. The method of claim 10 , further comprising adjusting a start frame of the steady-rate region to allow for an integer-to-integer mapping of digital input frames to digital output frames of the steady-rate region.