Prior art

This analysis identifies relevant prior art for US Patent 10218995, based on the citations provided in its Google Patents record (https://patents.google.com/patent/[US10218995](/patent/US10218995)/en). The focus is on patent documents cited as prior art. Due to the extensive list of cited patents, a representative selection of the most relevant prior art will be detailed, followed by a summary for other cited documents.

The core inventive concepts of US10218995, as understood from its definitions, involve a multi-layer video encoding system that leverages super-resolution processing to enhance encoding efficiency and quality. Key elements include:

• A first encoder processing standard resolution video.
• A first super-resolution enlarger and resolution converter processing the input video to create "super-resolution enlarged and converted pictures" at standard resolution, which contain enhanced frequency components.
• A second encoder using these enhanced pictures as encoding targets, and using both standard decoded pictures from the first encoder and super-resolution enlarged and converted decoded pictures (from a second super-resolution path) as reference pictures.
• The overall aim is hierarchical encoding, improving prediction by incorporating higher-frequency information obtained through super-resolution.

Most Relevant Prior Art for US10218995

Below are details for the most relevant prior art cited in US10218995, including full citations, dates, brief descriptions, and potential claims anticipated under 35 U.S.C. § 102.

1. US20030095713A1

• Full Citation: US 2003/0095713 A1 to Hiroshi Takagi et al., titled "Scalable video coding apparatus and method"
• Publication/Filing Date: Published May 22, 2003. Filed August 21, 2002.
• Brief Description: This patent application describes a scalable video coding apparatus and method that generates a base layer and an enhancement layer. The enhancement layer can use reconstructed images from the base layer for inter-layer prediction. It discusses using up-sampled base layer images as predictors for the enhancement layer. The invention aims to improve coding efficiency for scalable video.
• Potential Anticipation (35 U.S.C. § 102): US20030095713A1 potentially anticipates claims related to the general concept of scalable or hierarchical video coding where an enhancement layer utilizes a decoded and up-sampled version of a base layer for prediction. This is particularly relevant to the elements of US10218995 concerning a "first encoder" providing decoded pictures and a "second encoder" using reference pictures derived from these decoded pictures for inter-layer prediction (e.g., as described for the "first reference pictures" in the second encoder). However, it does not explicitly describe the "super-resolution enlargement" process with the aim of recreating lost frequency components or the use of super-resolution enlarged input pictures as encoding targets for the enhancement layer, which are key aspects of US10218995.

2. US20080031548A1

• Full Citation: US 2008/0031548 A1 to Shigeyuki Sakazawa et al., titled "Image coding system, image coding method, image decoding system, image decoding method, image coding program and image decoding program"
• Publication/Filing Date: Published February 7, 2008. Filed July 23, 2007.
• Brief Description: This application describes an image coding system that performs scalable coding, wherein a base layer and an enhancement layer are coded. The enhancement layer can utilize decoded images from the base layer, potentially after resolution conversion (e.g., upsampling), as reference for prediction. It focuses on reducing calculation amount and improving coding efficiency.
• Potential Anticipation (35 U.S.C. § 102): Similar to US20030095713A1, this reference potentially anticipates claims related to scalable video coding using inter-layer prediction from a decoded and resolution-converted (e.g., up-sampled) base layer. This would relate to claims in US10218995 involving the "first encoder" providing decoded pictures and the "second encoder" using reference pictures derived from these decoded pictures. The distinction for US10218995 still lies in the explicit super-resolution processing to reconstruct high-frequency information and the use of super-resolution processed input as encoding targets.

3. US6466701B1

• Full Citation: US 6,466,701 B1 to Takashi Morishige, titled "Image coding and decoding method and apparatus for improving image quality of decoded images"
• Publication/Filing Date: Granted October 15, 2002. Filed November 15, 2000.
• Brief Description: This patent describes an image coding and decoding method that encodes difference information between an input image and a predicted image, and then adds residual signal components to a decoded image to improve its quality, particularly for low bit rate coding. It mentions generating a higher-resolution image for prediction.
• Potential Anticipation (35 U.S.C. § 102): This patent touches upon image quality improvement and using higher-resolution images for prediction in a general sense. While it deals with improving decoded image quality, it doesn't appear to explicitly describe the specific multi-layered super-resolution encoding architecture of US10218995, particularly the generation of "super-resolution enlarged and converted pictures" from the input signal as an encoding target, or the use of two types of reference pictures (standard decoded vs. super-resolution decoded). Its relevance might be to broader concepts of using enhanced images for prediction.

4. US20070058912A1

• Full Citation: US 2007/0058912 A1 to Yoshikazu Ohno et al., titled "Image encoding method and apparatus, and image decoding method and apparatus"
• Publication/Filing Date: Published March 15, 2007. Filed September 12, 2006.
• Brief Description: This application describes an image encoding system that generates a high-resolution image from a low-resolution image using an interpolation process and then uses this high-resolution image for encoding. It discusses improving encoding efficiency for images of different resolutions.
• Potential Anticipation (35 U.S.C. § 102): This reference is highly relevant as it explicitly discusses generating a high-resolution image from a low-resolution image using interpolation and then using this high-resolution image for encoding. This directly relates to the concept of the "first super-resolution enlarger" (103) in US10218995 creating super-resolution enlarged pictures from standard resolution input. The distinction for US10218995 would be its specific multi-layer architecture, the subsequent resolution conversion back to standard resolution for the encoding target, and the dual reference picture mechanism (standard decoded vs. super-resolution decoded).

5. US20040170327A1

• Full Citation: US 2004/0170327 A1 to Mitsuyoshi Suzuki, titled "Video decoding method, video encoding method, and scalable video encoding/decoding system"
• Publication/Filing Date: Published September 2, 2004. Filed February 26, 2004.
• Brief Description: This patent application describes a scalable video encoding/decoding system that uses inter-layer prediction, specifically generating a prediction image for an enhancement layer from a decoded image of a base layer. It also covers the possibility of adapting the up-sampling filter based on image characteristics.
• Potential Anticipation (35 U.S.C. § 102): This reference, like others in scalable video coding, potentially anticipates claims of US10218995 related to using decoded base layer images, potentially up-sampled, for inter-layer prediction in an enhancement layer. The core difference remains US10218995's specific integration of a super-resolution enlargement process aimed at recovering frequency components for both input and decoded pictures, and the use of "super-resolution enlarged and converted pictures" as an encoding target.

---

Summary of Other Cited Prior Art Documents

Many other cited patent documents generally relate to various aspects of video coding, scalable video coding, motion estimation, deblocking filters, and resolution conversion techniques, which are foundational to video compression. While these contribute to the overall field, they may not individually anticipate the specific combination of super-resolution processing for both input and decoded video streams within a hierarchical encoding framework, where the super-resolution processed input also serves as an encoding target for an enhancement layer, as claimed in US10218995.

For instance, documents like:

• US7139332B2 (Kimura, "Scalable video coding apparatus and method") and US7072513B2 (Oki, "Multi-resolution image coding method and device") relate to scalable video coding or multi-resolution coding.
• Many others are likely directed to specific improvements in motion estimation, entropy coding, quantization, deblocking, or other standard components of video codecs (e.g., H.264/AVC, MPEG-4 SVC, JPEG2000). For a full 35 U.S.C. § 102 analysis, each of these would require a deep dive into their claims and specifications to determine if they disclose all elements of any specific claim of US10218995.

Given the explicit descriptions within US10218995 about leveraging "information on frequency components in the spatial direction and the temporal direction that has been potentially contained in the input moving pictures but unable to express to a sufficient degree by the standard resolution, to reconstruct on basis of a prescribed super-resolution process," the prior art most likely to be considered anticipatory would be those that detail similar super-resolution reconstruction within the context of a video encoding loop, particularly for scalable or hierarchical coding. The specific novelty of US10218995 seems to reside in the dual application of super-resolution (on original input and on decoded frames) and the use of the super-resolution enhanced input as an encoding target for an enhancement layer.