IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`Application No.
`Filed
`Confirmation No.
`
`16/591,903
`>
`: October 3, 2019
`:
`3237
`
`Examiner
`Art Unit
`
`: BOYLAN, JAMEST
`:
`2486
`
`Docket No.
`>
`735256.422C1
`First Named Inventor : Ryuichi KANOH
`Title
`- ENCODER, DECODER, ENCODING METHOD, AND DECODING
`METHOD
`
`INTERVIEW AGENDA
`
`Representative Claim:
`
`7.
`
`(Previously Presented) A decoder comprising:
`
`processing circuitry; and
`
`a memory coupled to the processing circuitry,
`
`wherein the processing circuitry is configuredto:
`
`select a first filter for a first block based at least on a prediction mode used for the
`
`first block;
`
`select a secondfilter for a second block; and
`
`change valuesofpixels in the first block and the second block by using thefirst
`
`filter and the secondfilter such that change amounts of the respective values are within
`
`respective clip widths, the pixels in the first block and the second block being arranged along a
`
`line across a boundary betweenthefirst block and the second block, and the clip widths are
`
`asymmetric with respect to the boundary.
`
`Support is found, for example, in 4][03 13 ]&[0342] of the application as published (US
`
`2020/0036975), reproduced below:
`
`[0313] When the calculated d is larger than the second threshold value (Yes in
`$104), loop filter 120 determinesa filter characteristic (S105), and performs
`
`

`

`Application No. 16/591,903
`Reply to Office Action Dated December 21, 2020
`
`filtering using the determinedfilter characteristic (S106). For example, a 5-tap
`filter of (1, 2, 2,2, 1)/81s used. Specifically, for p10 indicated in FIG.12, a
`calculation of (1 x p30+ 2 x p20 +2 x pl0+2 x q1l0+ 1 x q20)/8is
`performed. Here, in the filtering, clipping is performed so that variationfalls
`within a certain range without excessive smoothing. Clipping here is threshold
`processing which, for example, when a threshold value for clipping is tc and a
`pixel value to be filtered is q, only allowsthe filtered pixel value to take a
`
`value within the range of q +tc.
`
`BIG. 12
`
`e BLOCK BOUNDARY
`
`g
`
`Q
`
`=
`
`$
`
`®
`
`f
`>
`SAANN Aang SAN
`eofpodfbpsa|bpig , qgi®§ghfFgafgsQ
`
`
`AOLPAAASEROLLSIDSLLILSDLLORSYLSULDESAEDEDS
`LPOPLOLOLPAREDIDSSASUELEEEL
`
`heer
`
`
`pat
`pl
`wie.
`NAAANANNAAAAAAAAAS
`caeaaaaneneaenn)
`reo
`
`
`
`eiSfjpeefppeeiplefgis i ads
`xwhe
`
`
`
`ges asf ads
`3
`,
`
`gis
`8
`
`g
`
`3x
`
`pds
`
`eye
`pas
`
`wes
`~
`
`sis
`
`r
`
`[0342] Next, a fourth approach for changing weights asymmetricallyis
`described. Loop filter 120 performs a filter calculation using a filter
`coefficient of a reference filter. Next, when a change amount Ain pixel value
`
`before and after the Giter calculation exceeds a clin width which is a reference
`
`value, loop
`filter 120 clins the change amount Ato the clin width, Loop filter
`126 sets asymmetrical clin widths across a block boundary.
`
`
`
`
`In the invention as recited in claim 7, threshold processing is performed in that, where
`
`a clipping threshold is clip width tc and a pixel valueto be filtered is q, the filtered pixel value
`
`takes a value within the range of q+ tc. Further, asymmetrical clip widths are set across a
`
`block boundary.
`
`Rejection under 85 U.S.C. 108
`
`The office has made the following findingin the final office action:
`
`

`

`Application No. 16/591,903
`Reply to Office Action Dated December 21, 2020
`
`Glaims 7-8, 12-11, 114 and 16-17 are mgjected under 34 U.S.C. 103 as heing
`
`unpatentable over Narroschke et al. (herein after wit be rafarred to as Nerroschke} (US
`
`20140233859) in view of Norkin et al (herein after wil be referred to as Norkin} (US
`
`20130320814).
`
`«
`
`change values of pixels in the first block and the second block by using the first
`
`iRter and the second iter such fhat change amounts of the respective values are
`
`within respective clip widths, the clio wiclths are asymme§jiric wih respect fo the
`
`boundary.
`
`[See Narroschke [0196-41] Different clipping thresholds
`
`including Tel, To2 and Te3. Also, in 0138, these thresholds contro! the
`
`maximum and minimum clipping values. Also, see Fig. 5, Block boundary
`
`between block A and B.]
`
`However,
`
`in Narroschke,
`
`the first
`
`threshold value Tcl
`
`is not one of clipping
`
`thresholds (clip widths), and further, either the second threshold value Tc2 is used or the
`
`third threshold value Tc3 is used as a clip width. That is, clip widths Tc2 and Tc8 are never
`
`used together as two asymmetric clip widths across a block boundary. To the contrary, in
`
`Narroschke, either clip width of Te2 or Tc3 is symmetrically used across the block boundary.
`
`This is clear based on the disclosure of ¢§ [0133]-[0141] of Narroschke. Narroschke
`
`describes that the first threshold value Tcl is used to determine whether weakfiltering is to
`
`be applied, and not used as a clip width to clip variation amounts (910133]1&[0134]). When
`
`a weakfilter is to be applied, and p1 and/or q1 is furtherfiltered, the second threshold value
`
`Tc2 is to be used for clipping ({§[0136]&[0137]). On the other hand, whena strongfilteris
`to be applied, the third threshold value Tc3is to be used for clipping (410139]).
`
`Significantly, when a weak filter is applied, clip width Tc2 is used symmetrically
`
`across a block boundary and, when a strong filter is applied, clip width Tc3 is used
`
`symmetrically across a block boundary. Narroschke does not even suggest the possibility of
`
`using the two clip widths Tc2 and Tc3 at the same time asymmetrically across the block
`
`boundary.
`
`