Abstract: Rate-QP estimation for a B picture is disclosed which involves: providing an input group of pictures (GOP); selecting an input B picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input B picture. The outputting step may involve calculating intra/non-intra luma and chroma Rate-QP estimates from corresponding intra/non-intra luma and chroma histograms; offsetting the intra/non-intra chroma Rate-QP estimate to form respective offset intra/non-intra chroma estimates; and setting a bit rate corrected Rate-QP for the input B picture to a corrected sum of the previous estimates. The histograms are formed using an input of the lowest SATD forward, backward, and bidirectional prediction coefficients, and the intra prediction coefficients, where an intra/non-intra mode is selected, which results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in advanced video encoders.

Abstract: Rate-QP estimation for an I picture is disclosed which involves the steps of: providing an input group of pictures (GOP); selecting an input I picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input I picture. The outputting step may involve calculating intra luma and chroma Rate-QP estimates from corresponding intra luma and chroma histograms; offsetting the intra chroma Rate-QP estimate to form an offset intra chroma estimate; and setting a bit rate corrected Rate-QP for the input I picture to a corrected sum of the previous estimates. The histograms are formed with estimates of intra prediction coefficients, where an intra/non-intra mode is selected that results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in an advanced video encoder.

Abstract: Rate-QP estimation for a P picture is disclosed which involves the steps of: providing an input group of pictures (GOP); selecting an input P picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input P picture. The outputting step may involve calculating intra/non-intra luma and chroma Rate-QP estimates from corresponding intra/non-intra luma and chroma histograms; offsetting the intra/non-intra chroma Rate-QP estimate to form respective offset intra/non-intra chroma estimates; and setting a bit rate corrected Rate-QP for the input P picture to a corrected sum of the previous estimates. The histograms are formed with estimates of intra and forward prediction coefficients, where an intra/non-intra mode is selected that results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in an advanced video encoder.

Abstract: A method of intra mode prediction uses a block of pixels and their horizontal Hpos and vertical Vpos pixel positions and adjacent horizontal and vertical pixels within an input picture frame signal as inputs for a method for selecting a lowest Sum of Absolute Transformed Differences (SATD) intra mode among intra modes among the horizontal, vertical, and steady state (DC) intra modes for use in advanced video coding algorithms, such as MPEG-4 Part 10 and H.264/AVC. Associated costs are calculated for each of the intra modes are used to select the output of the best intra mode. The method reduces the unimproved computational cost of three 2D 4×4 Hadamard transformations (which is equivalent to 24 1D 4 point transformations) to just 4 1D 4 point transformations for a significant computational improvement. As horizontal and vertical panning is frequently used in video imagery, this improvement may reduce encoder processing by 80%.

Abstract: An apparatus and method are taught for estimating an optimized sub-pixel interpolation filter using iterative and non-iterative estimations as needed for sub-pixel motion compensation and motion estimation in a video codec for improving coding efficiency. Motion vector information and mode decisions are passed from the first encoding stage which uses predetermined interpolation to at least a second encoding stage which uses an estimated adaptive interpolation filter determined during the first encoding stage. Processing overhead is reduced within the subsequent stages. Embodiments are described in which additional stages perform iterative encoding and estimation of interpolation filter in an n-th iteration.

Abstract: The invention is an apparatus and method for estimating an optimized sub-pixel interpolation filter using iterative estimations as needed for sub-pixel motion compensation and motion estimation in a video codec for improving coding efficiency. Multiple iterations of adaptive interpolation filter estimation are performed including more than one iteration based on sub-pixel motion vectors. During testing of the inventive apparatus and method on various video segments, average bit rate reductions were exhibited of approximately 5%.

Abstract: Described herein are apparatus and methods for enhancing perceived visual quality of video which has been coded using enhanced intra predictive coding referred to as boundary adaptive intra predictions (BAIP). BAIP is performed in response to the type of prediction, and the smoothness of the boundary between a current block and its neighboring pixels which have already been coded. Intra prediction is performed separately for each pixel along the left side and/or top side within a current block, which takes into account the pixel itself and multiple pixels along the left and/or top border of that current block. The BAIP technique can be applied for different block sizes, including 4×4, 8×8, and 16×16, and includes vertical, horizontal, and DC intra predictions.

Abstract: An adaptive loop filter utilizing separable filters within a quadtree-based adaptive loop filter (QALF). The filters for at least the P and B pictures are replaced with a separable filter to provide Separable QALF (SQALF). In a preferred embodiment the filter comprises a Wiener filter as a best fit approximation of the non-separable filter. In response to decoding with the SQALF approach, computational complexity in the decoder is reduced without lowering perceived visual quality. Estimation of the separable filter is obtained in the encoder by iteratively minimizing the sum of squares error with the separable filter. The SQALF filtering is described within a video decoder.

Abstract: In one embodiment, a method for bypassing transform and quantization steps in a video compression is provided. The method determines one or more threshold values that are computed for a bypass of the transform and quantization steps. A sum of absolute transformed differences (SATD) value from a motion estimation of a signal being compressed using H.26x is received. The SATD value is compared to at least one of the one or more threshold values. It is then determined if the transform and quantization can be bypassed based on the comparison. For example, if the SATD value is less than the at least one of the one or more threshold values, then it may be determined that the coefficients of the transform are expected to be quantized to zero. Accordingly, the transform and quantization steps may be bypassed.

Abstract: A method and apparatus for performing quantization during video compression, such as within an MPEG-2 encoder. Inter-quantization is particularly recited in this implementation, although the teachings may be applied to intra-quantization as well. Computationally intensive portions of the quantization process are split into multiple components in response to differing execution probability. A first component is executed within each quantization iteration, and preferably is performed in parallel, such as using SIMD processing. At least one additional component, such as a second component, is serially processed in response to detecting a condition that does not arise with each loop iteration. Preferred embodiments of the invention replace division operations with a combination of multiply and shift operations, while making extended use of table lookup procedures to reduces processing overhead.

Abstract: A method and apparatus for performing quantization during video compression, such as within an MPEG-2 encoder. Intra-quantization is particularly recited, although the teachings may be applied to inter-quantization as well. Computationally intensive portions of the quantization process are split into multiple components in response to differing execution probability. A first component is executed within each quantization iteration, and preferably is performed in parallel, such as using SIMD processing. At least one additional component, such as a second component, is serially processed in response to detecting a condition that does not arise with each loop iteration. Preferred embodiments of the invention replace division operations with a combination of multiply and shift operations, while making extended use of table lookup procedures to reduces processing overhead.

Abstract: Rate-QP estimation for an I picture is disclosed which involves the steps of: providing an input group of pictures (GOP); selecting an input I picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input I picture. The outputting step may involve calculating intra luma and chroma Rate-QP estimates from corresponding intra luma and chroma histograms; offsetting the intra chroma Rate-QP estimate to form an offset intra chroma estimate; and setting a bit rate corrected Rate-QP for the input I picture to a corrected sum of the previous estimates. The histograms are formed with estimates of intra prediction coefficients, where an intra/non-intra mode is selected that results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in an advanced video encoder.

Abstract: Rate-QP estimation for a P picture is disclosed which involves the steps of: providing an input group of pictures (GOP); selecting an input P picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input P picture. The outputting step may involve calculating intra/non-intra luma and chroma Rate-QP estimates from corresponding intra/non-intra luma and chroma histograms; offsetting the intra/non-intra chroma Rate-QP estimate to form respective offset intra/non-intra chroma estimates; and setting a bit rate corrected Rate-QP for the input P picture to a corrected sum of the previous estimates. The histograms are formed with estimates of intra and forward prediction coefficients, where an intra/non-intra mode is selected that results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in an advanced video encoder.

Abstract: Rate-QP estimation for a B picture is disclosed which involves: providing an input group of pictures (GOP); selecting an input B picture within the GOP; and outputting, to a computer readable medium, a bit rate corrected Rate-QP, R(QP), for the input B picture. The outputting step may involve calculating intra/non-intra luma and chroma Rate-QP estimates from corresponding intra/non-intra luma and chroma histograms; offsetting the intra/non-intra chroma Rate-QP estimate to form respective offset intra/non-intra chroma estimates; and setting a bit rate corrected Rate-QP for the input B picture to a corrected sum of the previous estimates. The histograms are formed using an input of the lowest SATD forward, backward, and bidirectional prediction coefficients, and the intra prediction coefficients, where an intra/non-intra mode is selected, which results in a lowest SATD for each macroblock in the GOP. The methods may be implemented into a computer program, possibly resident in advanced video encoders.

Abstract: A method of intra mode prediction uses a block of pixels and their horizontal Hpos and vertical Vpos pixel positions and adjacent horizontal and vertical pixels within an input picture frame signal as inputs for a method for selecting a lowest Sum of Absolute Transformed Differences (SATD) intra mode among intra modes among the horizontal, vertical, and steady state (DC) intra modes for use in advanced video coding algorithms, such as MPEG-4 Part 10 and H.264/AVC. Associated costs are calculated for each of the intra modes are used to select the output of the best intra mode. The method reduces the unimproved computational cost of three 2D 4×4 Hadamard transformations (which is equivalent to 24 1D 4 point transformations) to just 4 1D 4 point transformations for a significant computational improvement. As horizontal and vertical panning is frequently used in video imagery, this improvement may reduce encoder processing by 80%.

Abstract: Picture level rate control systems, apparatuses and methods are described which indirectly control bit rate through peak signal to noise ratio (PSNR), thereby providing a stable PSNR while maintaining a given target average bit rate. In a preferred implementation, encoding is performed in conjunction with macroblock QP modulation. In this way, the picture level rate control provides stable temporal quality and the macroblock QP modulation provides stable spatial quality. By way of example, the number of bits generated in a group of pictures (GOP) is controlled by modulating the target bit allocation for a group of pictures (GOP) based on coded picture buffer (CPB) fullness. Macroblocks are encoded for each picture based on a rate-distortion model, which is modified in response to the relative fluctuation of PSNR of pictures within a group of pictures.

Abstract: The largest absolute value (LAV) is determined within a group of data. Based on the LAV, a difference table is identified that is to be used for selecting an optimal Huffman codebook for the group of data. The difference table is associated with two Huffman codebooks. Further, one or more indexes are calculated for the group of data using an expression associated with the two Huffman codebooks, and a size difference value is determined for the group of data using the calculated indexes and the difference table. Based on the determined size difference value, the optimal Huffman codebook is selected from the above two codebooks.

Abstract: In one embodiment, a method for bypassing transform and quantization steps in a video compression is provided. The method determines one or more threshold values that are computed for a bypass of the transform and quantization steps. A sum of absolute transformed differences (SATD) value from a motion estimation of a signal being compressed using H.26x is received. The SATD value is compared to at least one of the one or more threshold values. It is then determined if the transform and quantization can be bypassed based on the comparison. For example, if the SATD value is less than the at least one of the one or more threshold values, then it may be determined that the coefficients of the transform are expected to be quantized to zero. Accordingly, the transform and quantization steps may be bypassed.

Abstract: A method of generating a value for a missing pixel “x” by determining a “least harmful” local edge direction between pixels, or sub-pixels, on substantially opposing sides of the missing pixel, and interpolating the difference to arrive at a value for pixel “x”. The method involves generating sub-pixel values for locations within neighboring pixels, the sub-pixels may comprise half-pixels, quarter-pixels, three-quarter pixels, and so forth, wherein any fractional pixel quantity may be created. Absolute difference values are calculated between neighboring pixels, or sub-pixel values, to determine a least harmful local edge direction along which a value is generated for pixel “x” by interpolation.

Abstract: Methods for enhancing the performance of quantization operations by converting division operations to a combination of multiplication and shift operations, which are preferably performed on a processor supporting single-instruction multiple-data (SIMD) instructions. A table of mantissa and exponent values is created for a sufficient range of values for 1/a. During quantization, the mantissa and exponent values are found in the table 1/a for associated with a given quantization division operation given by b/a which is found according to the formula b/a=(b×A)>>n. Aspects are described for application to processors that do not support non-uniform shift operations, and for reducing the necessary bit-width of the operations to increase efficiency. The quantization method may be applied to protocols such as MPEG-2 and other similar formats.