Walter Gish - Oak Park CA, US Christopher Vogt - Laguna Niguel CA, US
Assignee:
Dolby Laboratories Licensing Corporation - San Francisco CA
International Classification:
H04N 7/36
US Classification:
37524016, 37524015
Abstract:
Optimal error metric function for motion estimation is determined and used for video coding and/or video processing of images. To do so, an initial motion estimation using an initial error metric function can be performed. This can produce motion prediction errors. If the initial error metric function is not the optimal error function, then a final motion estimation is performed using a selected optimal error metric function. In some embodiments, a shape of error distribution can be used to determine the optimal error metric function. Some example systems or devices for this motion estimation can include systems or devices for compression, temporal interpolation, and/or super-resolution processing.
Walter C. Gish - Oak Park CA, US Christopher J. Vogt - Laguna Niguel CA, US
Assignee:
Dolby Laboratories Licensing Corporation - San Francisco CA
International Classification:
H04N 7/12 H04N 11/02 H04N 11/04
US Classification:
37524003, 37524001, 37524002, 37524026
Abstract:
The quantization parameter QP is well-known in digital video compression as an indication of picture quality. Digital symbols representing a moving image are quantized with a quantizing step that is a function QSN of the quantization parameter QP, which function QSN has been normalized to the most significant bit of the bit depth of the digital symbols. As a result, the effect of a given QP is essentially independent of bit depth a particular QP value has a standard effect on image quality, regardless of bit depth. The invention is useful, for example, in encoding and decoding at different bit depths, to generate compatible, bitstreams having different bit depths, and to allow different bit depths for different components of a video signal by compressing each with the same fidelity (i. e. , the same QP).
The quantization parameter QP is well-known in digital video compression as an indication of picture quality. Digital symbols representing a moving image are quantized with a quantizing step that is a function QSof the quantization parameter QP, which function QShas been normalized to the most significant bit of the bit depth of the digital symbols. As a result, the effect of a given QP is essentially independent of bit depth—a particular QP value has a standard effect on image quality, regardless of bit depth. The invention is useful, for example, in encoding and decoding at different bit depths, to generate compatible, bitstreams having different bit depths, and to allow different bit depths for different components of a video signal by compressing each with the same fidelity (i. e. , the same QP).
Walter Gish - Oak Park CA, US Hyung-Suk Kim - Glendale CA, US
Assignee:
Dolby Laboratories Licensing Corporation - San Francisco CA
International Classification:
H04N 7/50
US Classification:
375240130, 375240120, 375E07211
Abstract:
Unbiased rounding of unsigned data is employed in the decoding or the encoding and decoding of digital bitstreams representing data-video when the video is encoded at a first bit depth and is decoded at a second bit depth, lower than the first bit depth. The unbiased rounding may be employed in processing that employs a prediction loop. When the data-compressed video is represented in frames, the unbiased rounding may be of inter-frame and/or intra-frame data.
Compatible Compression Of High Dynamic Range, Visual Dynamic Range, And Wide Color Gamut Video
Walter Gish - Oak Park CA, US Zhen Li - Cupertino CA, US Christopher Vogt - Laguna Niguel CA, US
Assignee:
DOLBY LABORATORIES LICENSING CORPORATION - San Francisco CA
International Classification:
H04N 7/26
US Classification:
37524025, 375E07027
Abstract:
A first image stream has a first dynamic range and a first color space. First and the second image streams are received in a layered codec. The second image stream has a second dynamic range, which is higher than the first dynamic range. The first image stream is in the codec's base layer; the second image stream is in its enhancement layer. The first image stream is encoded to obtain an encoded image stream, which is decoded to obtain a decoded image stream. The decoded image stream is converted from the first non-linear or linear color space to a second, different color space to obtain a color converted image stream. A higher dynamic range image representation of the color converted image stream is generated to obtain a transformed image stream. Inverse tone mapping parameters are generated based on the transformed image stream and the second image stream.
Walter Christian Gish - Oak Park CA, US Christopher J. Vogt - Laguna Niguel CA, US
Assignee:
DOLBY LABORATORIES LICENSING CORPORATION - San Francisco CA
International Classification:
H04N 7/26
US Classification:
37524003, 375E07139
Abstract:
The quantization parameter QP is well-known in digital video compression as an indication of picture quality. Digital symbols representing a moving image are quantized with a quantizing step that is a function QSN of the quantization parameter QP, which function QSN has been normalized to the most significant bit of the bit depth of the digital symbols. As a result, the effect of a given QP is essentially independent of bit depth a particular QP value has a standard effect on image quality, regardless of bit depth. The invention is useful, for example, in encoding and decoding at different bit depths, to generate compatible, bitstreams having different bit depths, and to allow different bit depths for different components of a video signal by compressing each with the same fidelity (i.e., the same QP).
WALTER C. GISH - OAK PARK CA, US SAMIR HULYALKAR - LOS GATOS CA, US
Assignee:
DOLBY LABORATORIES LICENSING CORPORATION - SAN FRANCISCO CA
International Classification:
H04N 7/30 H04N 7/50
US Classification:
37524016, 3752402, 375E07226, 375E07211
Abstract:
A visual dynamic range (VDR) signal and a standard dynamic range (SDR) signal are received. A first (e.g., MPEG-2) encoder encodes a base layer (BL) signal. A second encoder encodes an enhancement layer (EL). The EL signal represents information with which the VDR signal may be reconstructed, e.g., using the BL and the EL signals. The first encoder encodes the SDR signal with inverse discrete cosine transform (IDCT) coefficients that have a fixed precision, e.g., which represent fixed-point approximations of transform coefficients that may have arbitrary precisions. The BL signal is encoded in a stream that conforms with an Advanced Television Standards Committee (ATSC) standard. The EL is encoded in a stream that conforms with an ATSC enhanced vestigial sideband (E-VSB) standard. The BL and EL signals are combined; e.g., multiplexed, and transmitted together.
Quality Assessment Of Images With Extended Dynamic Range
DOLBY LABORATORIES LICENSING CORPORATION - San Francisco CA
International Classification:
G06K 9/68 G06K 9/34
US Classification:
382173, 382218
Abstract:
Given a set of two images with visual dynamic range (VDR), wherein the second image is an approximation of the first image, a standard dynamic range (SDR) equivalent peak value (EPV) is computed, wherein the EPV is smaller than the maximum pixel value of the first VDR image. An image assessment metric for the second image is computed using the EPV and an error measurement between the first image and the second image. The EPV may be computed using image statistics (e.g., mean and standard deviation) derived from the first VDR image or, if available, from its corresponding SDR image.
Resumes
Principal Member, Technical Staff At Dolby Laboratories