Obviousness

The provided patent document, US10237577, does not list specific prior art references (e.g., patent numbers, academic papers) in a dedicated "Prior Art section" that can be combined for an obviousness analysis. Instead, the patent's "Background Art" and "Description of Embodiments" sections describe the state of the art, primarily referencing the H.264/AVC video coding standard and general video compression techniques. Therefore, this analysis will proceed by combining general knowledge of the H.264/AVC standard, as described within US10237577, with common motivations of a person having ordinary skill in the art (PHOSITA) in the field of video encoding and decoding.

Legal Standard for Obviousness (35 U.S.C. § 103)

Under 35 U.S.C. § 103, a patent claim is obvious if "the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains." The Supreme Court in KSR Int'l Co. v. Teleflex Inc. emphasized that obviousness can be established by showing that a PHOSITA would have had a "motivation to combine" or "reason to modify" prior art elements with a reasonable expectation of success.

Analysis of US10237577 Claims

The core of US10237577 lies in applying "in-loop filtering" to a restored current block before storing it in a frame buffer for subsequent intra prediction. The patent asserts that this improves prediction accuracy and video compression efficiency. [Abstract, Summary of Invention]

Claim 1:
"An intra prediction method for video encoding and decoding, the method comprising steps of: generating a residual block of a current block through inverse quantization and inverse transformation; generating a prediction block of the current block through intra prediction; performing in-loop filtering on a current block obtained by summing the residual block and the prediction block step; and storing the current block subjected to the in-loop filtering in a frame buffer for intra prediction of a next encoding target block."

Hypothetical Prior Art Combination:

1. Prior Art 1 (H.264/AVC Standard):

   • Video Encoding/Decoding Fundamentals: The patent acknowledges that video data is typically "encoded by performing intra prediction or inter prediction, transformation, quantization, entropy coding, or the like on each picture of the video data in units of blocks." [Description] This includes "generating a residual block through inverse quantization and inverse transformation" as part of the reconstruction process within the encoder/decoder loop. [Description, S110]
   • Intra Prediction: The patent explicitly references H.264/AVC, stating, "in existing intra prediction of H.264/AVC, extrapolation is performed using nine modes." [Background Art, FIG. 2] This covers the step of "generating a prediction block of the current block through intra prediction."
   • Frame Buffer Usage: The concept of storing reconstructed blocks (or frames) in a "frame buffer for intra prediction of a next encoding target block" is a fundamental aspect of video coding standards like H.264/AVC, where previously encoded/decoded blocks serve as reference pixels for subsequent predictions. [Description, S140]
   • De-blocking Filtering: Crucially, the patent states, "In H.264 that is one existing coding method, there is de-blocking filtering that plays a similar role to filtering that is performed on left and up blocks encoded in the in-loop filtering step S 130 of the present invention." [Description, Second Embodiment] This indicates that de-blocking filtering was known in H.264/AVC and was applied to reconstructed blocks/frames to improve image quality and reduce artifacts like "block burn-in" at block boundaries.

2. Motivation to Combine (H.264/AVC with In-Loop De-blocking for Prediction):
   The patent itself identifies the problem that "the intra prediction has a drawback in that accurate prediction is difficult as a pixel to be predicted is away from a reference pixel." [Background Art] It further notes that "distortion between an original video and a restored video may occur or may degrade accuracy of subsequent intra prediction or inter-frame prediction." [Background Art]

   A person having ordinary skill in the art (PHOSITA) in video coding, seeking to address these known problems within the H.264/AVC framework, would have been motivated to combine the existing intra prediction and reconstruction processes with the known de-blocking filtering technique in an in-loop manner as claimed. The motivation stems from the understanding that:

   • Improved Reference Quality: If the quality of reference pixels used for prediction is improved, the accuracy of subsequent predictions should also improve.
   • De-blocking Purpose: De-blocking filters in H.264/AVC are specifically designed to reduce quantization errors and block artifacts, thereby improving the visual quality of reconstructed blocks.
   • Direct Solution to Stated Problem: By moving the de-blocking filter into the prediction loop to filter the "current block obtained by summing the residual block and the prediction block" (i.e., the reconstructed block) before it is stored in the reference frame buffer, the PHOSITA would directly provide higher-quality reference pixels for subsequent intra prediction. This would logically lead to "reduc[ing] a prediction error and increasing video compression efficiency by achieving more accurate prediction," as stated in the patent's objective. [Summary of Invention]

   The patent explicitly reinforces this motivation by stating that "a block in which block burn-in has been removed can be predicted by referencing a block in which the quantization error has been removed when the previously encoded block is referenced in a current frame in the intra prediction, which enables more excellent prediction." [Description, Second Embodiment] This statement articulates the direct causal link that would motivate a PHOSITA to apply such filtering in-loop.

   Therefore, combining the fundamental intra prediction, reconstruction, and frame buffering mechanisms of H.264/AVC with its known de-blocking filter, by placing the filter within the prediction loop to process reconstructed blocks before they are used as references, would have been an obvious modification to a PHOSITA seeking to improve prediction accuracy and reduce distortion, with a reasonable expectation of success.

Claim 7:
"The method of claim 1, wherein the step of performing in-loop filtering comprises performing the in-loop filtering on a down boundary and right boundary of the current block."

Obviousness of Claim 7 (in combination with Claim 1's obviousness):
Given the obviousness of applying in-loop filtering as in Claim 1, the specific application of filtering "on a down boundary and right boundary of the current block" would also be obvious to a PHOSITA. De-blocking filters in video coding standards often operate on block edges. [Description, First Embodiment, FIG. 6A and 6B] The processing order of blocks (e.g., left-to-right, top-to-bottom) naturally creates dependencies where the right and down boundaries of a current block become the left and up boundaries for adjacent, subsequently encoded/decoded blocks. Therefore, prioritizing filtering at these boundaries ensures that the most relevant reference pixels for future blocks are improved, making this a logical and obvious design choice for an in-loop filter in such a system. The patent itself notes, "filtering is performed on pixels placed at down and right block boundaries 511 and 512 of the encoded block in the prediction mode direction to reduce the quantization error." [Description, First Embodiment, FIG. 6A and 6B]

Claim 8:
"The method of claim 1, wherein the step of performing in-loop filtering comprises adjusting whether to perform filtering, the type of filtering, and intensity of the filtering based on an intra prediction direction of blocks adjacent to the current block."

Obviousness of Claim 8 (in combination with Claim 1's obviousness):
The concept of adaptive filtering is well-known in video coding. Existing de-blocking filters, including those in H.264/AVC, often adapt their strength or application based on various coding parameters, such as quantization parameters or block types, to optimize the balance between artifact reduction and detail preservation. [Description, Second Embodiment]
A PHOSITA, implementing an in-loop filter, would be motivated to make it adaptive to further enhance its effectiveness. Adjusting filtering parameters (presence, type, intensity) "based on an intra prediction direction of blocks adjacent to the current block" is a logical extension of adaptive filtering. For example, if adjacent blocks have a similar prediction direction, applying stronger filtering along that direction might be beneficial to reinforce continuity, as suggested by the patent. [Description, Second Embodiment, FIG. 7A] Conversely, if prediction directions are discontinuous, weaker or no filtering might be applied to avoid blurring essential details. [Description, Second Embodiment, FIG. 7B] Such adaptivity is a common optimization technique in video processing, and its application to the in-loop filtering of Claim 1 would be an obvious design choice for a PHOSITA aiming to achieve improved coding efficiency and subjective image quality. [Description, First Embodiment]