Back to results
US 9,287,852 B2Granted
Methods and systems for efficient filtering of digital signals
AssigneeGoogle Technology Holdings LLC
InventorChanchal Chatterjee
Life of this patent
Estimating the term…
Abstract
A method in a signal processor for filtering samples in a digital signal is provided. An approximate filtered sample is generated as a function of less than four samples of the digital signal. A correction is generated as a function of the less than four samples, and a filtered sample is generated by modifying the approximate filtered sample with the correction.
Claims13
A method for filtering in a signal processor, the method comprising:
- receiving as an input a digital signal;
- extracting a plurality of pixel values from the received digital signal;
- generating an approximate filtered pixel value as a function of three pixel values of the plurality of pixel values from the digital signal associated with a reference picture, the three pixel values including a first fixed-point pixel value, A, a second fixed-point pixel value, B, and a third fixed-point pixel value, C,
- the generating the approximate filtered pixel value including:
- calculating M=(B+C+1)>>1, and
- calculating the approximate filtered pixel value as X=(A+M+1)>>1,
- generating a correction as a function of the three pixel values;
- generating a filtered pixel value by modifying the approximate filtered pixel value with the correction;
- generating a predicted macroblock based on the filtered pixel value;
- and
- encoding a frame of a picture based on the predicted macroblock.
8 more elements
A method for filtering in a signal processor, the method comprising:
- receiving as an input a digital video signal;
- extracting three pixel values associated with a reference picture from the received digital video signal;
- generating an approximate filtered pixel value as a function of the three pixel values, the three pixel values including a first fixed-point pixel value, A, a second fixed-point pixel value, B, and a third fixed-point pixel value, C,
- the generating the approximate filtered pixel value including:
- calculating M=(B+C+1)>>1, and
- calculating the approximate filtered pixel value as X=(A+M+1)>>1,
- generating a correction as a function of the three pixel values;
- modifying the approximate filtered pixel value with the correction;
- generating a predicted macroblock based on the filtered pixel value;
- and
- encoding a frame of a picture based on the predicted macroblock.
8 more elements
Dependencies are inferred from the claim text.