Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Gregory Maxwell
On Wed, Apr 13, 2016 at 7:24 PM, Davinder Singh wrote: > good vectors? how can i improve them? since it search for every possible > place, it should give best match. can you give more details, why > surrounding vectors need to be considered? > > also i tried to compare it

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Michael Niedermayer
On Wed, Apr 13, 2016 at 10:30:35PM +0200, Michael Niedermayer wrote: > On Wed, Apr 13, 2016 at 10:01:35PM +0200, Michael Niedermayer wrote: > > On Wed, Apr 13, 2016 at 07:24:42PM +, Davinder Singh wrote: > > > good vectors? how can i improve them? since it search for every possible > > >

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Michael Niedermayer
On Wed, Apr 13, 2016 at 10:01:35PM +0200, Michael Niedermayer wrote: > On Wed, Apr 13, 2016 at 07:24:42PM +, Davinder Singh wrote: > > good vectors? how can i improve them? since it search for every possible > > place, it should give best match. can you give more details, why > > surrounding

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Thilo Borgmann
Am 13.04.16 um 22:01 schrieb Michael Niedermayer: > On Wed, Apr 13, 2016 at 07:24:42PM +, Davinder Singh wrote: >> good vectors? how can i improve them? since it search for every possible >> place, it should give best match. can you give more details, why >> surrounding vectors need to be

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Michael Niedermayer
On Wed, Apr 13, 2016 at 07:24:42PM +, Davinder Singh wrote: > good vectors? how can i improve them? since it search for every possible > place, it should give best match. can you give more details, why > surrounding vectors need to be considered? the goal is to find the true motion that is

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-13 Thread Davinder Singh
good vectors? how can i improve them? since it search for every possible place, it should give best match. can you give more details, why surrounding vectors need to be considered? also i tried to compare it with export_mvs, they seems quite similar. the export_mvs ones, sometimes use multiple

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-11 Thread Michael Niedermayer
On Sat, Apr 09, 2016 at 11:50:07PM +, Davinder Singh wrote: > ok, applied the changes. with new closest vector condition. will add > threshold now. it may be beyond the scope of this qualification task but to get good quality vectors the surrounding vectors will have to be considered in

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-10 Thread Davinder Singh
here are samples: big buck bunny: (block=16:search=16:step=2) http://www.mediafire.com/download/iy35rcr6d66733o/output1.mp4 matrixbench: (block=16:search=7) http://www.mediafire.com/download/ii3n9sn42bwp3sg/output_matrix.mp4 On Sun, Apr 10, 2016 at 5:20 AM Davinder Singh

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-09 Thread Davinder Singh
ok, applied the changes. with new closest vector condition. will add threshold now. On Sun, Apr 10, 2016 at 12:47 AM Paul B Mahol wrote: > On 4/9/16, Davinder Singh wrote: > > hi > > here's the new patch. > > > > changes: fixed bugs, added documentation >

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-09 Thread Paul B Mahol
On 4/9/16, Davinder Singh wrote: > hi > here's the new patch. > > changes: fixed bugs, added documentation > added a step option, by default, while searching for macro-block, it moves > by 1 px in search area. this can be increased now. > uses absolute values instead of

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-09 Thread Davinder Singh
hi here's the new patch. changes: fixed bugs, added documentation added a step option, by default, while searching for macro-block, it moves by 1 px in search area. this can be increased now. uses absolute values instead of squares. vector starts from center of macro-block, before it was at

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-06 Thread Paul B Mahol
> From 68560d7d4764587ad72f1901edc45450391f1378 Mon Sep 17 00:00:00 2001 > From: dsmudhar > Date: Mon, 4 Apr 2016 01:06:07 +0530 > Subject: [PATCH] motion estimation filter > > --- > libavfilter/Makefile | 1 + > libavfilter/allfilters.c | 1 + >

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-04 Thread Michael Niedermayer
On Sun, Apr 03, 2016 at 08:45:46PM +, Davinder Singh wrote: > sorry about that, > here is recreated patch. > > > DSM_ > > On Sun, Apr 3, 2016 at 11:09 PM Davinder Singh wrote: > > > Qualification task for Motion interpolation project. > > > > here is basic motion

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-03 Thread Davinder Singh
sorry about that, here is recreated patch. DSM_ On Sun, Apr 3, 2016 at 11:09 PM Davinder Singh wrote: > Qualification task for Motion interpolation project. > > here is basic motion estimation filter which uses block matching > technique, does full search in (default) 7

Re: [FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-03 Thread Michael Niedermayer
On Sun, Apr 03, 2016 at 05:39:51PM +, Davinder Singh wrote: > Qualification task for Motion interpolation project. > > here is basic motion estimation filter which uses block matching technique, > does full search in (default) 7 px region by 1px step. > > Thanks, > DSM_ >

[FFmpeg-devel] [Patch][GSoC] Motion Estimation filter

2016-04-03 Thread Davinder Singh
Qualification task for Motion interpolation project. here is basic motion estimation filter which uses block matching technique, does full search in (default) 7 px region by 1px step. Thanks, DSM_ 0001-added-motion-estimattion-filter.patch Description: Binary data