Prior art

US Patent 6950469, titled "Method for sub-pixel value interpolation," addresses techniques for generating sub-pixel values in digital video encoding and decoding. The patent aims to improve upon existing methods like TML5 and TML6, which suffered from issues such as increased memory usage, computational complexity, and reduced precision due to intermediate truncation and clipping operations. The core of US6950469's invention lies in specific interpolation schemes for various sub-pixel resolutions (e.g., half and quarter-pixel locations) that minimize these drawbacks, particularly by direct interpolation using weighted sums and diagonal averaging. The priority date for US6950469 is 2001-09-17.

Based on a review of the patent citations listed for US6950469 and their publication dates prior to its priority date, the following are identified as the most relevant prior art.

Most Relevant Prior Art for US6950469:

1. US5594813A

   • Full Citation: US5594813A, Fandrianto et al., "Method and apparatus for video compression using motion compensation with sub-pixel resolution".
   • Publication/Filing Date: Publication: 1997-01-14 (Filing: 1994-07-28).
   • Brief Description: This patent describes a method and apparatus for video compression that incorporates motion compensation with sub-pixel resolution. Such systems inherently require techniques for interpolating pixel values to generate sub-pixel values for more accurate motion estimation and compensation.
   • Potential Anticipation (35 U.S.C. § 102): This patent directly addresses video compression using motion compensation with sub-pixel resolution, anticipating the general field of the invention described in US6950469. It likely discloses methods for sub-pixel value interpolation that would broadly fall under the preamble of claim 1, which defines a method of interpolation to generate values for sub-pixels at fractional locations. Depending on the specific interpolation filters and averaging techniques employed in US5594813A, it could potentially anticipate the use of weighted sums or averages for generating sub-pixels, as generally described in claims 1(a), 1(b), or 1(c) of US6950469.

2. US5901248A

   • Full Citation: US5901248A, Fandrianto et al., "Method and apparatus for motion estimation and compensation with sub-pixel accuracy in video compression".
   • Publication/Filing Date: Publication: 1999-05-04 (Filing: 1997-01-14).
   • Brief Description: This patent details a method and apparatus for achieving sub-pixel accuracy in motion estimation and compensation for video compression. This necessitates the creation and use of sub-pixel values through interpolation to refine motion vectors beyond full-pixel resolution.
   • Potential Anticipation (35 U.S.C. § 102): Similar to US5594813A, this patent directly addresses the core problem that US6950469 aims to improve: sub-pixel accuracy in motion compensation for video. It would likely teach methods for interpolating sub-pixel values to achieve this accuracy. Its disclosure of specific interpolation methods, such as those employing weighted sums or averaging for different fractional pixel locations, could potentially anticipate elements of claim 1, particularly sub-pixel generation for motion vectors. If it describes distinct strategies for interpolating sub-pixels based on their horizontal, vertical, or diagonal fractional positions, it could further anticipate the specific approaches in claims 1(a), 1(b), or 1(c).

3. US6104753A

   • Full Citation: US6104753A, Kim et al., "Method and apparatus for motion compensated inter-frame coding using subpixel motion vectors".
   • Publication/Filing Date: Publication: 2000-08-15 (Filing: 1999-02-19).
   • Brief Description: This patent focuses on a method and apparatus for motion-compensated inter-frame coding that explicitly utilizes sub-pixel motion vectors. The use of sub-pixel motion vectors implies the necessity of interpolating image data to generate values at fractional pixel positions.
   • Potential Anticipation (35 U.S.C. § 102): This patent's title clearly indicates its relevance to the use of sub-pixel motion vectors in video coding, which requires sub-pixel interpolation. Therefore, it generally anticipates the need for such interpolation methods as stated in the preamble of claim 1 of US6950469. The specific techniques disclosed in US6104753A for calculating these sub-pixel values, especially if they involve direct interpolation of higher-resolution sub-pixels from lower-resolution ones or base pixels using weighted sums or averaging, would determine its potential anticipation of the particular steps outlined in claims 1(a), 1(b), and 1(c).

4. US6219464B1

   • Full Citation: US6219464B1, Greggain et al., "Motion compensated interpolation for video processing".
   • Publication/Filing Date: Publication: 2001-04-17 (Filing: 2000-06-21).
   • Brief Description: This patent describes methods and systems for motion-compensated interpolation specifically tailored for video processing. This direct focus on "motion compensated interpolation" suggests the presence of detailed techniques for generating intermediate pixel values for motion prediction.
   • Potential Anticipation (35 U.S.C. § 102): As this patent directly concerns "motion compensated interpolation for video processing," it is highly relevant. It anticipates the general concept of performing interpolation within video coding to support motion compensation, as laid out in the preamble of claim 1. Its specific methods for determining sub-pixel values, particularly if they involve weighted averaging, multi-stage interpolation schemes, or handling of different fractional resolutions, could potentially anticipate the interpolation steps defined in claims 1(a), 1(b), and 1(c) of US6950469.

5. US5568597A

   • Full Citation: US5568597A, Nakayama et al., "Image interpolation method and apparatus".
   • Publication/Filing Date: Publication: 1996-10-22 (Filing: 1995-02-21).
   • Brief Description: This patent discloses an image interpolation method and apparatus, indicating its relevance to the fundamental process of generating intermediate pixel values to increase image resolution or for other image processing tasks.
   • Potential Anticipation (35 U.S.C. § 102): While not exclusively tied to motion compensation, this patent directly addresses "image interpolation," which is a core component of US6950469. It generally anticipates the process of generating sub-pixel values from existing pixels. Depending on the detailed interpolation algorithms it teaches—such as the use of weighted sums or averages of neighboring pixels for fractional locations, or specific multi-step interpolation for different resolutions—it could potentially anticipate the methods described in claims 1(a), 1(b), or 1(c), especially concerning the mathematical operations for deriving new pixel values.

Other Relevant Prior Art (published before 2001-09-17):

• GB2205707A: Sony Corp, "Television standards converters," published 1988-12-14. This patent likely involves general image interpolation techniques for converting between different television standards. It broadly anticipates the use of interpolation in image processing but may not detail sub-pixel interpolation methods with the specificity of US6950469's claims.
• US4816913A: Harney et al., "Video compression system for digital transmission of still and motion images," published 1989-03-28. This patent introduces general concepts of video compression and motion compensation but may not delve into the specific sub-pixel interpolation methodologies of US6950469.
• EP0573290A2: Publication: 1993-12-08. Without a specific title or detailed description, its exact relevance to the claims of US6950469 is difficult to ascertain, though it is likely related to video or image processing.
• US5485279A: Yonemitsu et al., "Image data processing apparatus," published 1996-01-16. This patent broadly covers image data processing, which could include interpolation, but the specific techniques relevant to sub-pixel value interpolation of US6950469 are not immediately apparent from the title alone.
• US5521642A: Park, "Television signal processing method and apparatus for improving image quality," published 1996-05-28. Improving image quality often involves interpolation, so this patent may cover general interpolation techniques, potentially anticipating broad aspects of generating interpolated pixel values.
• US5570436A: Fukushima et al., "Image processing apparatus and method for scaling an image," published 1996-10-29. Scaling images typically requires interpolation, suggesting that this patent might contain general interpolation methods that could broadly anticipate aspects of US6950469's claims related to generating values at fractional locations.
• US6252576B1: Nottingham, "Method and apparatus for generating a scaled image," published 2001-06-26. Similar to US5570436A, this patent concerns image scaling, implying interpolation techniques. Its detailed interpolation methods would need to be reviewed to determine specific anticipation of the detailed steps in claims 1(a), 1(b), or 1(c).