X-Git-Url: https://bilbo.iut-bm.univ-fcomte.fr/and/gitweb/book_gpu.git/blobdiff_plain/8ad1643f80bdf5681bdb9cca04ff30378cb44cb8..c56162cd0df6cb604942d765ca5430092d4199dd:/BookGPU/Chapters/chapter3/ch3.aux

diff --git a/BookGPU/Chapters/chapter3/ch3.aux b/BookGPU/Chapters/chapter3/ch3.aux
index 0d1505e..b6d8bfd 100644
--- a/BookGPU/Chapters/chapter3/ch3.aux
+++ b/BookGPU/Chapters/chapter3/ch3.aux
@@ -23,6 +23,7 @@
 \@writefile{toc}{\contentsline {chapter}{\numberline {4}Implementing a fast median filter}{29}}
 \@writefile{lof}{\addvspace {10\p@ }}
 \@writefile{lot}{\addvspace {10\p@ }}
+\@writefile{toc}{\author{Gilles Perrot}{}}
 \@writefile{toc}{\contentsline {section}{\numberline {4.1}Introduction}{29}}
 \@writefile{toc}{\contentsline {section}{\numberline {4.2}Median filtering}{30}}
 \@writefile{toc}{\contentsline {subsection}{\numberline {4.2.1}Basic principles}{30}}
@@ -61,68 +62,71 @@
 \newlabel{lst:kernelMedian3RegTri9}{{4.2}{36}}
 \@writefile{lol}{\contentsline {lstlisting}{\numberline {4.2}3$\times $3 median filter kernel using one register per neighborhood pixel and bubble sort}{36}}
 \@writefile{toc}{\contentsline {subsection}{\numberline {4.4.2}Further optimization}{36}}
-\@writefile{lof}{\contentsline {figure}{\numberline {4.4}{\ignorespaces Comparison of pixel throughputs on GPU C2070 and CPU for generic median, in 3$\times $3 median register-only and \textit  {libJacket}.\relax }}{37}}
+\@writefile{toc}{\contentsline {subsubsection}{\numberline {4.4.2.1}Reducing register count }{36}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.4}{\ignorespaces Comparison of pixel throughputs on GPU C2070 and CPU for generic median, 3$\times $3 median register-only and \textit  {libJacket}.\relax }}{37}}
 \newlabel{fig:compMedians1}{{4.4}{37}}
 \@writefile{lof}{\contentsline {figure}{\numberline {4.5}{\ignorespaces Forgetful selection with the minimal element register count. Illustration for 3$\times $3 pixel window represented in a row and supposed sorted.\relax }}{37}}
 \newlabel{fig:forgetful_selection}{{4.5}{37}}
-\@writefile{toc}{\contentsline {subsubsection}{\numberline {4.4.2.1}Reducing register count}{37}}
-\newlabel{lst:medianForget1pix3}{{4.3}{38}}
-\@writefile{lol}{\contentsline {lstlisting}{\numberline {4.3}3$\times $3 median filter kernel using the minimum register count of 6 to find the median value by forgetful selection method}{38}}
-\@writefile{toc}{\contentsline {subsubsection}{\numberline {4.4.2.2}More data output per thread}{39}}
-\@writefile{lof}{\contentsline {figure}{\numberline {4.6}{\ignorespaces Illustration of how window overlapping is used to combine 2 pixel selections in a 3$\times $3 median kernel.\relax }}{40}}
-\newlabel{fig:median3_overlap}{{4.6}{40}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.6}{\ignorespaces Determination of the Median value by the forgetful selection process, applied to a $3\times 3$ neighborhood window.\relax }}{38}}
+\newlabel{fig:forgetful3}{{4.6}{38}}
+\newlabel{lst:medianForget1pix3}{{4.3}{39}}
+\@writefile{lol}{\contentsline {lstlisting}{\numberline {4.3}3$\times $3 median filter kernel using the minimum register count of 6 to find the median value by forgetful selection method. The optimal thread block size is 128 on GTX280 and 256 on C2070.}{39}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.7}{\ignorespaces Illustration of how window overlapping is used to combine 2 pixel selections in a 3$\times $3 median kernel.\relax }}{40}}
+\newlabel{fig:median3_overlap}{{4.7}{40}}
+\@writefile{toc}{\contentsline {subsubsection}{\numberline {4.4.2.2}More data output per thread}{40}}
 \newlabel{lst:medianForget2pix3}{{4.4}{40}}
 \@writefile{lol}{\contentsline {lstlisting}{\numberline {4.4}3$\times $3 median filter kernel processing 2 output pixel values per thread using combined forgetful selection.}{40}}
-\@writefile{lof}{\contentsline {figure}{\numberline {4.7}{\ignorespaces Comparison of pixel throughput on GPU C2070 for the different 3$\times $3 median kernels.\relax }}{41}}
-\newlabel{fig:compMedians2}{{4.7}{41}}
-\@writefile{toc}{\contentsline {section}{\numberline {4.5}A 5$\times $5 and more median filter }{41}}
-\@writefile{lof}{\contentsline {figure}{\numberline {4.8}{\ignorespaces Reducing register count in a 5$\times $5 register-only median kernel outputting 2 pixels simultaneously. The first 7 forgetful selection stages are common to both processed center pixels. Only the last 5 selections have to be done separately.\relax }}{42}}
-\newlabel{fig:median5overlap}{{4.8}{42}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.8}{\ignorespaces Comparison of pixel throughput on GPU C2070 for the different 3$\times $3 median kernels.\relax }}{41}}
+\newlabel{fig:compMedians2}{{4.8}{41}}
+\@writefile{toc}{\contentsline {section}{\numberline {4.5}A 5$\times $5 and more median filter }{42}}
 \newlabel{sec:median5}{{4.5.1}{42}}
 \@writefile{toc}{\contentsline {subsection}{\numberline {4.5.1}A register-only 5$\times $5 median filter }{42}}
 \newlabel{lst:medianForget2pix5}{{4.5}{42}}
 \@writefile{lol}{\contentsline {lstlisting}{\numberline {4.5}kernel 5$\times $5 median filter processing 2 output pixel values per thread by a combined forgetfull selection.}{42}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.9}{\ignorespaces Reducing register count in a 5$\times $5 register-only median kernel outputting 2 pixels simultaneously. The first 7 forgetful selection stages are common to both processed center pixels. Only the last 5 selections have to be done separately.\relax }}{43}}
+\newlabel{fig:median5overlap}{{4.9}{43}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.10}{\ignorespaces First iteration of the $5\times 5$ selection process, with $k_{25}=14$, which shows how Instruction Level Parallelism is maximized by the use of an incomplete sorting network. Arrows represent the result of the swapping function, with the lowest value at the starting point and the highest value at the end point.\relax }}{43}}
+\newlabel{fig:median5overlap}{{4.10}{43}}
 \@writefile{lot}{\contentsline {table}{\numberline {4.2}{\ignorespaces Performance of various 5$\times $5 median kernel implementations, applied on 4096$\times $4096 pixel image with C2070 GPU card.\relax }}{44}}
 \newlabel{tab:median5comp}{{4.2}{44}}
-\@writefile{toc}{\contentsline {subsection}{\numberline {4.5.2}Fast approximated n$\times $n median filter }{44}}
-\@writefile{lot}{\contentsline {table}{\numberline {4.3}{\ignorespaces Measured performance of one generic pseudo-separable median kernel applied to 4096$\times $4096 pixel image with various window sizes.\relax }}{45}}
-\newlabel{tab:medianSeparable}{{4.3}{45}}
+\@writefile{toc}{\contentsline {subsection}{\numberline {4.5.2}Fast approximated n$\times $n median filter }{45}}
 \newlabel{lst:medianSeparable}{{4.6}{45}}
 \@writefile{lol}{\contentsline {lstlisting}{\numberline {4.6}generic pseudo median kernel.}{45}}
-\newlabel{img:sap_example_ref}{{4.9(a)}{46}}
+\newlabel{img:sap_example_ref}{{4.11(a)}{46}}
 \newlabel{sub@img:sap_example_ref}{{(a)}{46}}
-\newlabel{img:sap_example_sep_med3}{{4.9(b)}{46}}
+\newlabel{img:sap_example_sep_med3}{{4.11(b)}{46}}
 \newlabel{sub@img:sap_example_sep_med3}{{(b)}{46}}
-\newlabel{img:sap_example_sep_med5}{{4.9(c)}{46}}
+\newlabel{img:sap_example_sep_med5}{{4.11(c)}{46}}
 \newlabel{sub@img:sap_example_sep_med5}{{(c)}{46}}
-\newlabel{img:sap_example_sep_med3_it2}{{4.9(d)}{46}}
+\newlabel{img:sap_example_sep_med3_it2}{{4.11(d)}{46}}
 \newlabel{sub@img:sap_example_sep_med3_it2}{{(d)}{46}}
-\@writefile{lof}{\contentsline {figure}{\numberline {4.9}{\ignorespaces Exemple of separable median filtering (smoother), applied to salt \& pepper noise reduction.\relax }}{46}}
+\@writefile{lof}{\contentsline {figure}{\numberline {4.11}{\ignorespaces Exemple of separable median filtering (smoother), applied to salt \& pepper noise reduction.\relax }}{46}}
 \@writefile{lof}{\contentsline {subfigure}{\numberline{(a)}{\ignorespaces {Airplane image, corrupted with by salt and pepper noise of density 0.25}}}{46}}
 \@writefile{lof}{\contentsline {subfigure}{\numberline{(b)}{\ignorespaces {Image denoised by a $3\times 3$ separable smoother}}}{46}}
 \@writefile{lof}{\contentsline {subfigure}{\numberline{(c)}{\ignorespaces {Image denoised by a $5\times 5$ separable smoother}}}{46}}
 \@writefile{lof}{\contentsline {subfigure}{\numberline{(d)}{\ignorespaces {Image background estimation by a $55\times 55$ separable smoother}}}{46}}
-\newlabel{fig:sap_examples2}{{4.9}{46}}
-\@writefile{toc}{\contentsline {section}{Bibliography}{47}}
+\newlabel{fig:sap_examples2}{{4.11}{46}}
+\@writefile{lot}{\contentsline {table}{\numberline {4.3}{\ignorespaces Measured performance of one generic pseudo-separable median kernel applied to 4096$\times $4096 pixel image with various window sizes.\relax }}{47}}
+\newlabel{tab:medianSeparable}{{4.3}{47}}
 \@setckpt{Chapters/chapter3/ch3}{
 \setcounter{page}{49}
 \setcounter{equation}{0}
 \setcounter{enumi}{3}
 \setcounter{enumii}{0}
 \setcounter{enumiii}{0}
-\setcounter{enumiv}{9}
+\setcounter{enumiv}{0}
 \setcounter{footnote}{0}
 \setcounter{mpfootnote}{0}
-\setcounter{part}{1}
+\setcounter{part}{2}
 \setcounter{chapter}{4}
 \setcounter{section}{5}
 \setcounter{subsection}{2}
 \setcounter{subsubsection}{0}
 \setcounter{paragraph}{0}
 \setcounter{subparagraph}{0}
-\setcounter{figure}{9}
+\setcounter{figure}{11}
 \setcounter{table}{3}
-\setcounter{numauthors}{0}
+\setcounter{numauthors}{1}
 \setcounter{parentequation}{0}
 \setcounter{subfigure}{0}
 \setcounter{lofdepth}{1}
@@ -134,6 +138,13 @@
 \setcounter{algocfline}{2}
 \setcounter{algocfproc}{2}
 \setcounter{algocf}{2}
+\setcounter{nprt@mantissa@digitsbefore}{0}
+\setcounter{nprt@mantissa@digitsafter}{0}
+\setcounter{nprt@exponent@digitsbefore}{0}
+\setcounter{nprt@exponent@digitsafter}{0}
+\setcounter{nprt@digitsfirstblock}{0}
+\setcounter{nprt@blockcnt}{0}
+\setcounter{nprt@cntprint}{0}
 \setcounter{proposition}{0}
 \setcounter{theorem}{0}
 \setcounter{exercise}{0}