Dear,<br><br>What i mean was,,,First I calculated integer level best motion vector using SAD as error criterion. that is the one (block) which gives minimum SAD . In this blk size is 16 x 16 and reference frame is previous frame(non interpolated)
<br><br>Then surrounding best integer level motion vector I choosed the eight half pixel locations and searched for the besxt match among this using SAD as the criterion. In this case also blk size was 16 x16. Now I compared the minimum SAD in both cases and if half pixel min SAD is best then I will go for quarter pixel level....In half and quarter pixel level the blk size was 16 x16 but the reference frame was previous frame interpolated upto quarter pixel level.
<br><br>Ex:<br><br>curr blk to be matched (2x2) <br> <br>100 102<br><br>90 95<br><br>Reference (non interpolated):<br><br>80 83 95 100<br><br>100 40 60 75<br><br>120 60 80 90<br><br>90 50 70 88
<br><br><br>Reference (non interpolated) used as reference for integer pixel ME.<br>
<br>
80 83 95 100<br>
<br>
100 40 60 75<br>
<br>
120 60 80 90<br>
<br>
90 50 70 88<br><br>Reference (interpolated) frame used as reference for half and quarter pixel ME.<br>
<br> I Q H Q I Q H Q I <br>
80 <span style="font-weight: bold;">82</span> <span style="text-decoration: underline;">81</span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>82</span> 83 <span style="font-weight: bold;">
86</span> <span style="text-decoration: underline;">89</span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>85</span> 95 <span style="font-weight: bold;">93</span> <span style="text-decoration: underline;">
97</span> <span style="font-weight: bold;">98</span> 100<br>
<br>
100 <span style="font-weight: bold;">85</span> <span style="text-decoration: underline;">70</span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>55</span> 40 <span style="font-weight: bold;">
45 </span> <span style="text-decoration: underline;">50</span> <span style="font-weight: bold;">55</span> 60 <span style="font-weight: bold;"> 64</span> <span style="text-decoration: underline;">68 </span> <span style="font-weight: bold;">
72</span> 75<br>
<br>
120 <span style="font-weight: bold;">105</span> <span style="text-decoration: underline;">90</span> <span style="font-weight: bold;">75</span> 60 <span style="font-weight: bold;">65</span> <span style="text-decoration: underline;">
70</span> <span style="font-weight: bold;">75</span> 80 <span style="font-weight: bold;">83</span> <span style="text-decoration: underline;">84</span><span style="font-weight: bold;"> </span> <span style="font-weight: bold;">
88 </span> 90<br>
<br>
90 <span style="font-weight: bold;">60</span> <span style="text-decoration: underline;">70</span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>40</span> 50 <span style="font-weight: bold;">
55</span> <span style="text-decoration: underline;">60 </span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>65</span> 70 <span style="font-weight: bold;">75</span> <span style="text-decoration: underline;">
79</span> <span style="font-weight: bold;"><span style="font-weight: bold;"></span>84</span> 88<br><br>This is how i implemented ....<br><br>Regards,<br>Maheshwar.<br><br><br><div><span class="gmail_quote">On 3/29/07,
<b class="gmail_sendername">Radek Czyz</b> <<a href="mailto:radoslaw@syskin.cjb.net">radoslaw@syskin.cjb.net</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Hi, can you explain more clearly? It's not ME that gives you SAD - it's<br>SAD that can be used to perform ME. The other way around ~<br><br>mahesh k wrote:<br>> Ok, thx for ur reply.....Can u plz tell me whether the quarter pixel ME
<br>> will give the min SAD for all the macroblocks within a frame.? Also plz<br>> tell me where I can find the test results regarding this?<br>><br>> regards,<br>> mahesh<br>><br>> On 3/28/07, *Radek Czyz* <
<a href="mailto:radoslaw@syskin.cjb.net">radoslaw@syskin.cjb.net</a><br>> <mailto:<a href="mailto:radoslaw@syskin.cjb.net">radoslaw@syskin.cjb.net</a>>> wrote:<br>><br>> > U plz tell me whether this was correct or not.
<br>><br>> Well yeah, it's a good (although wasteful in terms of memory) way of<br>> doing it.<br>><br>> You ask for "correct" but ME is never defined, so all methods are<br>> "correct". Just some are more efficient (give better or faster result)
<br>> than others.<br>><br>> mahesh k wrote:<br>> > I went through the JM reference code but I was not able to understand<br>> > it. Recently i implemented quarter pixel motion estimation. First I
<br>> > interpolated the whole reference frame upto quarter pixel level and<br>> > stored this in a buffer. Then I found the integer pixel motion<br>> vector<br>> > using the previos frame as reference frame (non interpolated
<br>> reference<br>> > frame) and the search criterion i used was SAD. Then I performed<br>> the sub<br>> > pixel refinement using the best integer pixel motion vector<br>> coordinates
<br>> > as center using the quarterly interpolated frame as reference.<br>> ><br>> > U plz tell me whether this was correct or not.<br>> ><br>> > Regards,<br>> > Mahesh.
<br>> ><br>> > On 3/28/07, *Tomas Carnecky* <<a href="mailto:tom@dbservice.com">tom@dbservice.com</a><br>> <mailto:<a href="mailto:tom@dbservice.com">tom@dbservice.com</a>><br>> > <mailto:
<a href="mailto:tom@dbservice.com">tom@dbservice.com</a> <mailto:<a href="mailto:tom@dbservice.com">tom@dbservice.com</a>>>> wrote:<br>> ><br>> > mahesh k wrote:<br>> > > I didnot get u exactly..Can u tell me in detail...
<br>> > ><br>> ><br>> > RTFM means 'Read The Fucking Manual', see[1]. He replaced the<br>> 'M' with a<br>> > 'S' which he afterwards explained to be, quote, 'sourcecode or
<br>> > standard'. That means, he was trying to tell you 'Read The F***<br>> > sourcecode' or Read The F*** standard'.<br>> > IOW (you can use wikipedia, too, to look up what that
<br>> means), he told<br>> > you to read the sourcecode (which is freely available) or the<br>> H.264<br>> > standard. The information you're looking for is in both<br>> locations,
<br>> > hence, quote, 'your choice'.<br>> ><br>> > ><br>> > > On 3/27/07, *Loren Merritt* <<a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu
</a><br>> <mailto:<a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu</a>><br>> > <mailto: <a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu</a><br>> <mailto:
<a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu</a>>><br>> > > <mailto: <a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu</a><br>> <mailto:<a href="mailto:lorenm@u.washington.edu">
lorenm@u.washington.edu</a>><br>> > <mailto:<a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu</a><br>> <mailto:<a href="mailto:lorenm@u.washington.edu">lorenm@u.washington.edu
</a>> >>> wrote:<br>> > ><br>> > > On Tue, 27 Mar 2007, mahesh k wrote:<br>> > ><br>> > > > Can any one plz tell me how to choose the refernece
<br>> frame for<br>> > > > integer level and quarter pixel motion estimation.<br>> > ><br>> > > RTFS. (sourcecode or standard, your choice)<br>> > >
<br>> ><br>> > tom<br>> ><br>> > [1] <a href="http://en.wikipedia.org/wiki/RTFM">http://en.wikipedia.org/wiki/RTFM</a><br>> <<a href="http://en.wikipedia.org/wiki/RTFM">
http://en.wikipedia.org/wiki/RTFM</a>><br>> ><br>> > --<br>> > This is the x264-devel mailing-list<br>> > To unsubscribe, go to: <a href="http://developers.videolan.org/lists.html">
http://developers.videolan.org/lists.html</a><br>> ><br>> ><br>><br>> --<br>> This is the x264-devel mailing-list<br>> To unsubscribe, go to: <a href="http://developers.videolan.org/lists.html">
http://developers.videolan.org/lists.html</a><br>><br>><br><br>--<br>This is the x264-devel mailing-list<br>To unsubscribe, go to: <a href="http://developers.videolan.org/lists.html">http://developers.videolan.org/lists.html
</a><br><br></blockquote></div><br>