<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic ippiFilterGaussianBorder in Intel® Integrated Performance Primitives</title>
    <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168857#M26871</link>
    <description>&lt;P&gt;The&amp;nbsp;ippiFilterGaussianBorder puzzles me, notably the sigma parameter. Compared to doing the Gaussian Blur with a&amp;nbsp;matrix,&amp;nbsp;ippiFilterGaussianBorder gives me what looks like unprecise results. This is with a typical kernel size 15 or 31. Of course, a&amp;nbsp;Gaussian Blur is separable, so we can replace the matrix with two one-dimensional matrices.&lt;/P&gt;&lt;P&gt;Can you give an example of a matrix or two separable matrices&amp;nbsp;where&amp;nbsp;ippiFilterGaussianBorder is (or should be) equivalent to&amp;nbsp;ippiFilterBorder ?&lt;/P&gt;&lt;P&gt;Regards,&lt;/P&gt;&lt;P&gt;Adriaan van Os&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
    <pubDate>Mon, 03 Dec 2018 09:13:56 GMT</pubDate>
    <dc:creator>Adriaan_van_Os</dc:creator>
    <dc:date>2018-12-03T09:13:56Z</dc:date>
    <item>
      <title>ippiFilterGaussianBorder</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168857#M26871</link>
      <description>&lt;P&gt;The&amp;nbsp;ippiFilterGaussianBorder puzzles me, notably the sigma parameter. Compared to doing the Gaussian Blur with a&amp;nbsp;matrix,&amp;nbsp;ippiFilterGaussianBorder gives me what looks like unprecise results. This is with a typical kernel size 15 or 31. Of course, a&amp;nbsp;Gaussian Blur is separable, so we can replace the matrix with two one-dimensional matrices.&lt;/P&gt;&lt;P&gt;Can you give an example of a matrix or two separable matrices&amp;nbsp;where&amp;nbsp;ippiFilterGaussianBorder is (or should be) equivalent to&amp;nbsp;ippiFilterBorder ?&lt;/P&gt;&lt;P&gt;Regards,&lt;/P&gt;&lt;P&gt;Adriaan van Os&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Mon, 03 Dec 2018 09:13:56 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168857#M26871</guid>
      <dc:creator>Adriaan_van_Os</dc:creator>
      <dc:date>2018-12-03T09:13:56Z</dc:date>
    </item>
    <item>
      <title>Asking above question again.</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168858#M26872</link>
      <description>&lt;P&gt;Asking above question again. Either my understanding of&amp;nbsp;ippiFilterGaussianBorder is wrong, or it has some bugs.&amp;nbsp;&lt;/P&gt;&lt;P&gt;Can you give an example of a matrix or two separable matrices&amp;nbsp;where&amp;nbsp;ippiFilterGaussianBorder is (or should be) equivalent to&amp;nbsp;ippiFilterBorder ?&lt;/P&gt;&lt;P&gt;Regards,&lt;/P&gt;&lt;P&gt;Adriaan&amp;nbsp;van Os&lt;/P&gt;</description>
      <pubDate>Mon, 28 Jan 2019 08:58:54 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168858#M26872</guid>
      <dc:creator>Adriaan_van_Os</dc:creator>
      <dc:date>2019-01-28T08:58:54Z</dc:date>
    </item>
    <item>
      <title>Hi Adriaan van Os,</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168859#M26873</link>
      <description>&lt;P&gt;Hi Adriaan van Os,&lt;/P&gt;&lt;P&gt;for kernelSize = 15 and sigma = .5.5 coefficients are the next:&lt;/P&gt;&lt;P&gt;k0 =&amp;nbsp;0.038978182&lt;/P&gt;&lt;P&gt;k1 = 0.048321572&lt;/P&gt;&lt;P&gt;k2 = 0.057956705&lt;/P&gt;&lt;P&gt;k3 = 0.067252662&lt;/P&gt;&lt;P&gt;k4 = 0.075501995&lt;/P&gt;&lt;P&gt;k5 = 0.082006926&lt;/P&gt;&lt;P&gt;k6 = 0.086175894&lt;/P&gt;&lt;P&gt;k7 = 0.087612126&lt;/P&gt;&lt;P&gt;k8&amp;nbsp;= 0.086175894&lt;/P&gt;&lt;P&gt;k9&amp;nbsp;= 0.082006926&lt;/P&gt;&lt;P&gt;k10&amp;nbsp;= 0.075501995&lt;/P&gt;&lt;P&gt;k11&amp;nbsp;= 0.067252662&lt;/P&gt;&lt;P&gt;k12 = 0.057956705&lt;/P&gt;&lt;P&gt;k13&amp;nbsp;= 0.048321572&lt;/P&gt;&lt;P&gt;k14&amp;nbsp;=&amp;nbsp;0.038978182&lt;/P&gt;&lt;P&gt;the filter itself is separable - therefore row/column coefficients are the same. The only 1 fact that is not reflected in the documentation - that these coefficients are normalized on their initial sum (coeff are obtained according to the formula in the manual and then divided by their sum).&lt;/P&gt;&lt;P&gt;regards, Igor&lt;/P&gt;</description>
      <pubDate>Mon, 28 Jan 2019 14:08:49 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168859#M26873</guid>
      <dc:creator>Igor_A_Intel</dc:creator>
      <dc:date>2019-01-28T14:08:49Z</dc:date>
    </item>
    <item>
      <title>Thanks.</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168860#M26874</link>
      <description>&lt;P&gt;Thanks.&lt;/P&gt;&lt;P&gt;What sigma is ".5.5" ?&lt;/P&gt;&lt;P&gt;Regards,&lt;/P&gt;&lt;P&gt;Adriaan van Os&lt;/P&gt;</description>
      <pubDate>Wed, 30 Jan 2019 12:05:07 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168860#M26874</guid>
      <dc:creator>Adriaan_van_Os</dc:creator>
      <dc:date>2019-01-30T12:05:07Z</dc:date>
    </item>
    <item>
      <title>Hi,</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168861#M26875</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;this one (in bold):&lt;/P&gt;&lt;P&gt;IPPAPI(IppStatus, ippiFilterGaussianInit,(IppiSize roiSize, Ipp32u kernelSize, &lt;STRONG&gt;Ipp32f sigma&lt;/STRONG&gt;,&lt;BR /&gt;&amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; IppiBorderType borderType, IppDataType dataType, int numChannels,&lt;BR /&gt;&amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; &amp;nbsp; IppFilterGaussianSpec* pSpec, Ipp8u* pBuffer))&lt;BR /&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;regards, Igor&lt;/P&gt;</description>
      <pubDate>Wed, 30 Jan 2019 13:43:37 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/ippiFilterGaussianBorder/m-p/1168861#M26875</guid>
      <dc:creator>Igor_A_Intel</dc:creator>
      <dc:date>2019-01-30T13:43:37Z</dc:date>
    </item>
  </channel>
</rss>

