<div dir="ltr"><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">Hi Simon,</font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">sorry for the mistake, not"<span style="font-size:14px;font-family:arial,sans-serif">the backprojection should take much longer time than</span></font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff"><span style="font-size:14px;font-family:arial,sans-serif">sart algorithm</span>" , but "<span style="font-size:14px;font-family:arial,sans-serif">the backprojection should take much longer time than</span></font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff"><span style="font-size:14px;font-family:arial,sans-serif">forward projection in sart algorithm</span>".</font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff"><br></font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">BTW, how many cores does your computer have?? M</font><span style="color:rgb(0,0,255)">ine is 24 cores.</span></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">is it can explain the reason why it takes much longer time on my computer than yours?</font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">Regards</font></div><div class="gmail_default" style="font-family:'trebuchet ms',sans-serif"><font color="#0000ff">Guangming</font></div></div><div class="gmail_extra"><br clear="all"><div><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><font style="line-height:23.32400131225586px" face="arial, helvetica, sans-serif" color="#674ea7"><b><i>Guangming Zang (Alex)</i></b></font><div style="line-height:23.32400131225586px"><font face="arial, helvetica, sans-serif"><b><i><font color="#0b5394">King Abdullah University of Science and Technology</font><font color="#073763">(KAUST)</font></i></b></font></div><div style="line-height:23.32400131225586px"><span style="font-family:arial;font-size:13px;line-height:20.020000457763672px"><font color="#0b5394"><b><i>University of Chinese Academy of Sciences(UCAS)</i></b></font></span><font face="arial, helvetica, sans-serif"><b><i><font color="#073763"><br></font></i></b></font></div><div style="line-height:23.32400131225586px"><span style="font-family:arial;font-size:13px;line-height:20.020000457763672px"><font color="#0b5394"><b><i><br></i></b></font></span></div><div style="line-height:23.32400131225586px"><span style="font-family:arial;font-size:13px;line-height:20.020000457763672px"><font color="#0b5394"><b><i><br></i></b></font></span></div><div style="line-height:23.32400131225586px"><br></div></div></div></div></div></div>
<br><div class="gmail_quote">2015-07-27 15:28 GMT+03:00 Simon Rit <span dir="ltr"><<a href="mailto:simon.rit@creatis.insa-lyon.fr" target="_blank">simon.rit@creatis.insa-lyon.fr</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">I can try. Forward and back projection have the same algorithmic<br>
complexity but voxel based backprojection benefits from several<br>
optimizations in terms of memory management and computations that<br>
makes it faster. The best is to look at the code for further<br>
details... although both have far from being optimal. And I did not<br>
get the sentence "the backprojection should take much longer time than<br>
sart algorithm." because SART contains a backprojection and other<br>
steps so SART is obviously longer than the bp alone.<br>
Your log is strange and I don't see what steps are not timed that<br>
would take most of the time. I did the same test on my computer and<br>
here is my result:<br>
<br>
OS-SR-466:tmp srit$ rtksimulatedgeometry -n 360 -o geo<br>
OS-SR-466:tmp srit$ rtkprojectshepploganphantom -g geo -o proj.mha<br>
--dimension 512<br>
OS-SR-466:tmp srit$ rtksart -p . -r proj.mha -g geo -o SART_SL3.mha<br>
<span class=""> -f Joseph -b VoxelBasedBackProjection --newspacing 0.5 --dimension<br>
128,128,128 --spacing 1,1,1 --origin -64,-64,-64 -l 0.5 -n 3 --time<br>
1<br>
</span>Recording elapsed time...<br>
SARTConeBeamReconstructionFilter timing:<br>
Extraction of projection sub-stacks: 0.288571 s<br>
Multiplication by zero: 0.131672 s<br>
Forward projection: 34.3612 s<br>
Subtraction: 0.203409 s<br>
Multiplication by lambda: 0.146459 s<br>
Ray box intersection: 1.30755 s<br>
Division: 0.187294 s<br>
Multiplication by the gating weights: 0 s<br>
Displaced detector: 0.278408 s<br>
Back projection: 11.8456 s<br>
Volume update: 0 s<br>
It took... 53.2765 s<br>
<br>
Simon<br>
<span class=""><br>
On Mon, Jul 27, 2015 at 10:41 AM, Guangming Zang<br>
<<a href="mailto:guangming.zang@kaust.edu.sa">guangming.zang@kaust.edu.sa</a>> wrote:<br>
> Hi Simon,<br>
> Thanks for your reply.<br>
> would you pls help and explain why backprojection is expected to take<br>
> shorter time than forward projection?? because i was thinking if no caching<br>
> step, the backprojection should take much longer time than sart algorithm.<br>
</span>> yes, i run rtksart for 2 times <a href="http://once.it" rel="noreferrer" target="_blank">once.it</a> took 12xxs, similar to the time<br>
<span class="im HOEnZb">> consumed of 3 times's sart, which much slower than my own application.<br>
> BTW, D drive is local disk drive, and 360pro_SL_Vol128_512.mha are 360<br>
> shapp-logan projections(512*512 resolution each)<br>
> rtksart -p . -r 360pro_SL_Vol128_512.mha -g geometry.xml -o<br>
> ../Result_SL512/SART_SL3.mha -f Joseph -b VoxelBasedBackProjection<br>
> --newspacing 0.5 --dimension 128,128,128 --spacing 1,1,1 --origin<br>
> -64,-64,-64 -l 0.5 -n 3 --time 1<br>
><br>
> and i will try reader->Update() like what you said.<br>
> Thanks<br>
> Guangming<br>
><br>
><br>
><br>
> Guangming Zang (Alex)<br>
> King Abdullah University of Science and Technology(KAUST)<br>
> University of Chinese Academy of Sciences(UCAS)<br>
><br>
><br>
><br>
><br>
</span><span class="im HOEnZb">> 2015-07-27 8:59 GMT+03:00 Simon Rit <<a href="mailto:simon.rit@creatis.insa-lyon.fr">simon.rit@creatis.insa-lyon.fr</a>>:<br>
>><br>
>> Hi Guangming,<br>
>> It's not surprising to me that the backprojection is faster than the<br>
>> forward projection, that's what I expect. If the total time is longer,<br>
>> that's probably that some individual steps are not included in the total<br>
>> time. Can you try to add<br>
>> reader->Update();<br>
>> before the line<br>
>><br>
>> itk::TimeProbe totalTimeProbe;<br>
>><br>
>> in rtksart.cxx? It may be that all the reading operations are done but not<br>
>> timed in the sart->Update(). Why they are so long, I don't know, is your D:<br>
>> drive a network drive? Do you observe the same behavior if you do rtksart 2<br>
>> times in a row?<br>
>><br>
>> Simon<br>
>><br>
>><br>
>><br>
>> On Mon, Jul 27, 2015 at 12:30 AM, Guangming Zang<br>
</span><span class="im HOEnZb">>> <<a href="mailto:guangming.zang@kaust.edu.sa">guangming.zang@kaust.edu.sa</a>> wrote:<br>
>>><br>
>>> Hi RTK community,<br>
>>> i am using SART algorithm to reconstruct an object.<br>
>>> But in this new RTK version, the time recording seems a little weird:<br>
>>> the total time is 1219.12s , but adding the time cost in different<br>
>>> stages is not 1291.12 s. especially for "backprojection" part, only 16.6051s<br>
>>> to reconstruct a 128^3 volume ?? even shorter than forward projection part.<br>
>>> BTW, the -f and -b are Joseph and VoxelBasedBackProjection, respectively,<br>
>>> both multi-threading i think.<br>
>>> Can anyone tell me what's going on?<br>
>>> Thanks<br>
>>> Regards<br>
>>> Guangming<br>
>>><br>
>>><br>
</span><div class="HOEnZb"><div class="h5">>>> Guangming Zang (Alex)<br>
>>> King Abdullah University of Science and Technology(KAUST)<br>
>>> University of Chinese Academy of Sciences(UCAS)<br>
>>><br>
>>><br>
>>><br>
>>><br>
>>> ________________________________<br>
>>> This message and its contents, including attachments are intended solely<br>
>>> for the original recipient. If you are not the intended recipient or have<br>
>>> received this message in error, please notify me immediately and delete this<br>
>>> message from your computer system. Any unauthorized use or distribution is<br>
>>> prohibited. Please consider the environment before printing this email.<br>
>><br>
>><br>
><br>
><br>
> ________________________________<br>
> This message and its contents, including attachments are intended solely for<br>
> the original recipient. If you are not the intended recipient or have<br>
> received this message in error, please notify me immediately and delete this<br>
> message from your computer system. Any unauthorized use or distribution is<br>
> prohibited. Please consider the environment before printing this email.<br>
</div></div></blockquote></div><br></div>
<br>
<div><hr></div><font face="Arial" size="1">This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.</font>