Abstract: A method of speeding the computation of backward motion vector estimates which are derived from fields which are available from forward motion vector estimates as based on nearest neighbor criterion. The method allows for the generation of multi-frame backwards motion vectors. The invention of the present method for estimating backward motion vectors is particularly well-suited for use with the MPEG-2 video standard.

Abstract: A method of and system for encoding a sequence of pictures without overflowing or underflowing a decoder buffer is disclosed. In one embodiment, the method uses a single-pass algorithm to store picture data that has been encoded using an MPEG standard. The method is used to monitor a characteristic of a macroblock, such as the type of picture encoded within the macroblock (e.g., I-picture, B-picture, and P-picture). A storage parameter based on the type of picture is dynamically generated and used to store the macroblock. Storage parameters include any combination of bits allocation, quantization step, or other parameters that ensure that the decoder buffer neither overflows nor underflows. Storage parameters are determined for the current picture, on-the-fly, and are thus optimized. A decoder buffer is well suited for providing smooth pictures during scene cuts and scene changes.

Abstract: A method and system is provided for calculating motion vectors of macroblocks in a digital image of a digital video stream. The method and system reduces the computational overhead of calculating motion vectors computing difference measures using a predetermined pattern of pixels in each macroblock rather than all the pixels of the macroblock. Reduction of computational overhead can be further enhanced first using a subpattern, i.e., a sub-sample of the predetermined pattern on a subset of the macroblocks to determine close matching macroblocks and then using the predetermined pattern to determine the best matching macroblock from within the close matching macroblocks.

Abstract: A method and system is provided for calculating motion vectors of macroblocks in a digital image of a digital video stream. The method and system reduces the computational overhead of calculating motion vectors computing difference measures using a multi-phase computational scheme. Specifically, the pixel blocks the previous image are divided into different groups. The closest matching pixel block of each group is determined in a first phase. Then a more accurate difference measure is used to determine the origin block from among the closest matching pixel blocks.

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: A method and system is provided for calculating motion vectors of macroblocks in a digital image of a digital video stream. The method and system reduces the computational overhead of calculating motion vectors by limiting the search for the origin block to a coarse search window and a fine search window within the coarse search window. The difference measure is computed for only a subset of pixel blocks within the coarse search window to reduce the computational overhead. However, to increase accuracy, the difference measure of all the pixel blocks in the fine search window are computed. The pixel block having the smallest difference measure is selected as the origin block.

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.

Abstract: A method of speeding the computation of backward motion vector estimates which are derived from fields which are available from forward motion vector estimates as based on nearest neighbor criterion. The method allows for the generation of multi-frame backwards motion vectors. The invention of the present method for estimating backward motion vectors is particularly well-suited for use with the MPEG-2 video standard.

Abstract: The present invention provides an apparatus and method for adaptive entropy encoding/decoding in a multimedia compression system. For encoding, the steps include: A) parsing a predetermined number of quantized transform coefficients into a plurality of coefficient groups in a predetermined scanning order and converting the coefficient groups into a plurality of parameter sets according to a predetermined scheme and storing the parameter sets wherein, where a last coefficient group comprises all zero quantized transform coefficients, the last coefficient group is discarded; B) sending, in accordance with a signal from the encoder controller, a current parameter set of the plurality of parameter sets in the predetermined scanning order; C) adaptively selecting a current entropy encoder based on a state parameter of a last parameter set; D) encoding a current parameter set to provide entropy-encoded information bits. Decoding is accomplished in a fashion that mirrors the encoding method.

Abstract: The present invention provides an apparatus and method for, in at least one of: encoding and decoding, scanning transform coefficients in an image/video encoded/decoder. In encoding, the steps include: A) segregating a predetermined number of quantized transform coefficients into a plurality of coefficient zones and storing the coefficient zones in a memory unit; B) examining each coefficient to determine which zone/zones has/have at least one non-zero coefficient; C) scanning each coefficient of the zones having at least one non-zero coefficient in a predetermined scanning order; D) encoding which coefficient zone contains at least one non-zero coefficient to provide first entropy-encoded information bits; E) determining whether at least one coefficient zone is to be encoded; F) encoding the non-zero coefficients to provide second entropy encoded information bits. Decoding is accomplished in mirror fashion.

Abstract: The present invention provides a method (200) and an apparatus (100) for spatially adaptive filtering for video encoding. The apparatus filters a video sequence prior the encoding process. The apparatus comprises a noise variance determiner (102), a local variance determiner (104), a noise visibility function determiner (106), a Gaussian kernel determiner (108), and a convolver (110). The apparatus removes noise directly from a Displaced Frame Difference, DFD, signal. This novel approach removes noise and miscellaneous high frequency components from the DFD signal without the introduction of the filtering artifacts characteristic of current techniques. By reducing the miscellaneous high frequency components, the present invention is capable of reducing the amount of information that must be encoded by the video encoder without substantially degrading the decoded video sequence.

Abstract: The present invention is method, rate controller, and system for preventing overflow and underflow of a decoder buffer in a video compression system. First, a plurality of quantized video bits is received from an encoder (102). The plurality of quantized video bits correspond to a video frame type. Then, a virtual buffer is created, in a rate controller, to model fullness of a decoder buffer, based on the quantized video bits, to produce a virtual buffer fullness (104). A quantization stepsize estimate is finally determined, in a rate controller, based on at least a first predetermined target bit allocation, the video frame type, and the virtual buffer fullness (106). The quantization step size estimate will prevent the overflow and underflow of a decoder buffer.

Abstract: The present invention is method and apparatus for preventing overflow and underflow of an encoder buffer in a video compression system. A virtual buffer is created in a rate controller to model the decoder buffer fullness (102). A sequence of bits is generated by an encoder (104). The encoder is controlled by the rate controller to prevent a decoder buffer underflow and overflow. Then, the sequence of bits is received by the encoder buffer to produce a bitstream (106). The bitstream corresponds to an instantaneous channel bitrate. The bitstream is transmitted from the encoder buffer to a decoder buffer following a delay (108). The delay is controlled by a rate controller to synchronize an encoder buffer fullness with a virtual buffer fullness (110). The synchronization prevents overflow and underflow of the encoder buffer.

Abstract: Efficient and adaptable motion estimation between flames in a video sequence is achieved in a video compression system 100 by first performing a series of comparisons in a first region between a present video block of a present video frame and a comparison video block of a previously stored video frame based on a first search point displacement pattern 500. Then, similar comparisons are performed in a second and third region based on a second and third search point displacement pattern respectively 501, 502. A best match video block is determined as a result of the series of comparisons 503. The series of comparisons may be made for a predetermined number of video blocks 600 and video statistics may be collected for the comparisons 601. If the video statistics are unfavorable 602 or if a predetermined time has expired 603, then the region used for searching for a best match block may be changed.

Abstract: The present invention provides a method (100) and device (200) for allocating bits to video pictures in a video compression system. A sequence of video frames or pictures is received, and the number of bits used to compress a previously encoded frame and a signal-to-noise ratio for the immediately previously encoded video frame are determined. Rate distortion model parameters are updated. Then, first, second, and third deviations are determined. Finally, the bit allocation for the frame is determined based on the first number of bits used to compress a previously encoded frame, the first deviation, the second deviation, and the third deviation.

Abstract: Whether to encode a present video block of an incoming video frame, i.e. intra code, or whether to encode the difference between the present video block of the incoming frame and a best match video block from a previously stored video frame, i.e. non-intra code, is determined in a video compression system 100. First the present video block is divided into a predetermined number of sub blocks 301. Then an average pixel value is calculated for each sub block 302 and used along with the individual pixel values of the sub block to determine sub block deviations 303. The sub block deviations are used to compute a present video block deviation 303. The present video block deviation is then compared with an error deviation 306 that is calculated based upon the present video block pixel values and the best match video block pixel values 305. If the comparison is favorable, then the difference between the present video block and the best match video block are encoded.

Abstract: The present invention is a method (100) and apparatus (300) for adaptive entropy encoding/decoding of a plurality of quantised transform coefficients in a video/image compression system. For encoding, first, a predetermined number of quantized transform coefficients are received in a predetermined order, giving a generally decreasing average power. Then the quantized transform coefficients are parsed into a plurality of coefficient groups. When the last coefficient group comprises all zero quantized coefficients, it is discarded. The coefficient groups are then converted into a plurality of parameter sets in the predetermined order. A current parameter set is obtained from the parameter sets in the reverse order of the predetermined order. A current entropy encoder is selected adaptively based on the previously selected entropy encoder and the previous parameter set. The current parameter set is encoded by the current entropy encoder to provide entropy encoded information bits.