This file is a concatenation of two HTML files that John Walker wrote for the
original 'pnmshadow' program. It is not all relevant to the current
'ppmshadow' program.
<html>
<head>
<title>pnmshadow: How it Works</title>
</head>
<body>
<center>
<h1><img src="figures/how_title.jpg" width=417 height=116 alt="pnmshadow: How it Works"></h1>
</center>
<hr>
<p>
This document describes the process, including PBMplus commands
and the intermediate images they create, by
which <b><a href="./">pnmshadow</a></b>
adds black shadows to source images.
A <a href="how-t.html">companion document</a>
describes how translucent shadows are created when the
<b>-t</b> option is specified.
<h3>The Starting Point</h3>
Let's start with the following source image, 536 pixels wide and 141
pixels high. We convert the image from whatever form in which
it was originally created (GIF, JPEG, etc.) to a PPM file before
processing it with <b>pnmshadow</b>.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadin.gif" width=536 height=141 alt="Input image">
</table>
</center>
<h3>The Blank Background</h3>
We start by determining the size of the input image with
<b>pnmfile</b> and then constructing an image with the same
size as the input image consisting entirely of the background
color, which is defined as the color of the pixel at the upper
left corner of the source image. This is performed by the
command:
<p>
<pre>
pnmcut 0 0 1 1 <em>ifile</em> | pnmscale -xsize <em>xsize</em> -ysize <em>ysize</em> ><em>fname</em>-5.ppm
</pre>
<p>
yielding the image:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt5.gif" width=536 height=141 alt="Blank background image">
</table>
</center>
<h3>The Positive Mask</h3>
A positive mask image is created in which all pixels of the background
color are set to white and all other pixels are black. This is accomplished
by subtracting the blank background image from the input (using the
<tt>-difference</tt> option on <b>pnmarith</b> to avoid clipping at
zero or the maximum pixel value), then inverting the result and
thresholding it to a monochrome bitmap.
<p>
<pre>
pnmarith -difference <em>ifile</em> <em>fname</em>-5.ppm | pnminvert | ppmtopgm | pgmtopbm -thresh -value 1.0 ><em>fname</em>-1.ppm
</pre>
<p>
This produces the following mask image.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt1.gif" width=536 height=141 alt="Positive mask image">
</table>
</center>
<h3>The Blurred Image</h3>
Since we wish to simulate a shadow from a nearby extended
light source rather than a sharp shadow as cast by the
Sun, we need to prepare a blurred version of the original
image. If the <b>-t</b> option is not specified on
<b>pnmshadow</b> the shadow cast by an object of any color
is always black, so the positive mask serves as the source image
when preparing the shadow. A convolution kernel which averages
the number of pixels specified by the <b>-b</b> option
(default 11), written into the temporary file <tt><em>fname</em>-2.ppm</tt>
in ASCII PGM format, and then the blurred image is created with
the command:
<p>
<pre>
pnmconvol <em>fname</em>-2.ppm <em>fname</em>-1.ppm ><em>fname</em>-3.ppm
</pre>
<p>
With the default blur setting of 11 pixels, the blurred image
below is generated.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt3.jpg" width=536 height=141 alt="Blurred shadow">
</table>
</center>
<h3>Shadow on Background Color</h3>
Having generated the blurred shadow from the monochrome mask image,
it will consist of pixels ranging from white to black on a white
background. In order to preserve the background color in the
original image, we multiply the shadow by the blank background color
image created previously. White pixels take on the background color
and pixels belonging to the shadow are scaled to be relative to the
background.
<p>
<pre>
pnmarith -multiply <em>fname</em>-3.ppm <em>fname</em>-5.ppm ><em>fname</em>-10.ppm
</pre>
<p>
This yields the following shadow, with the background of the
original image.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt10.jpg" width=536 height=141 alt="Shadow with background color">
</table>
</center>
<h3>Offset Shadow Clip</h3>
Shadows, even bogus ones like we're generating, usually look best when
cast by a light source diagonally displaced from the centre of the
shadow-casting object. To achieve this effect, we first cut a
rectangle from the blurred shadow image reduced in size by the
the number of pixels specified by the <b>-b</b> option
which default to half the blur (<b>-b</b>) setting. The
<em>xsize</em> and <em>ysize</em> arguments in the following
command are the size of the input image in pixels less the shadow
displacement in the respective axis.
<p>
<pre>
pnmcut 0 0 <em>xsize</em> <em>ysize</em> <em>fname</em>-10.ppm ><em>fname</em>-4.ppm
</pre>
<p>
The shadow clip is the identical to the shadow on background color, but
smaller by the offset in each direction.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt4.jpg" width=531 height=136 alt="Offset shadow clip">
</table>
</center>
<h3>Offset Shadow</h3>
Now we're ready to assemble the shadow offset by the specified number
of pixels. We do this by pasting the image cut in the previous step
into the blank background, yielding an image the same size as the
source image with the blurred shadow displaced to the right and
down.
<p>
<pre>
pnmpaste -replace <em>fname</em>-4.ppm <em>xoffset</em> <em>yoffset</em> <em>fname</em>-5.ppm ><em>fname</em>-6.ppm
</pre>
<p>
This gives the following result:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt6.jpg" width=536 height=141 alt="Offset shadow">
</table>
</center>
<h3>Inverse Mask</h3>
In order to stitch everything together, we need an inverse of the
mask prepared earlier--one where black pixels represent the background
and all other material is white. This is easily accomplished by
running the positive mask through <b>pnminvert</b>:
<p>
<pre>
pnminvert <em>fname</em>-1.ppm ><em>fname</em>-7.ppm
</pre>
<p>
yielding:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt7.gif" width=536 height=141 alt="Inverse mask">
</table>
</center>
<h3>Masked Input Image</h3>
Now we use the inverse mask prepared in the previous step to create
an image containing all non-background pixels from the source image,
with background pixels set to black. We simply multiply the
inverse mask by the source image:
<p>
<pre>
pnmarith -multiply <em>ifile</em> <em>fname</em>-7.ppm ><em>fname</em>-8.ppm
</pre>
<p>
<em>et voilà:</em>
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt8.gif" width=536 height=141 alt="Masked Input Image">
</table>
</center>
<h3>Shadow with Source Masked</h3>
Our last intermediate step before joining the image with its
shadow is preparing a shadow image with all non-background pixels
in the source image set to black. This ensures that when we add
the image and the shadow, the shadow will not override any pixel
in the source image.
<p>
<pre>
pnmarith -multiply <em>fname</em>-6.ppm <em>fname</em>-1.ppm ><em>fname</em>-9.ppm
</pre>
<p>
This is accomplished by multiplying the shadow by the positive
mask image, which sets all non-background pixels in the source
image to black:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt9.jpg" width=536 height=141 alt="Shadow with Source Masked">
</table>
</center>
<h3>The Final Product</h3>
At long last, we're ready to put together the pieces and deliver
the result to our ever-patient user. This amounts simply to
adding the masked input image (consisting solely of non-background
pixels from the original image) to the shadow with source masked
(in which all source pixels are black):
<p>
<pre>
pnmarith -add <em>fname</em>-8.ppm <em>fname</em>-9.ppm
</pre>
<p>
The resulting image, with shadow, is as follows:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadout.jpg" width=536 height=141 alt="Output: Image with shadow">
</table>
</center>
<h3>Smooth Operator</h3>
Since many computer graphics programs create sharp edges on
text, it's often best to create an image at a greater resolution
than that used for presentation, then scale it to the final
resolution with a tool which resamples the image, thus
minimising jagged edges by averaging adjacent
pixels. Using the output of <b>pnmshadow</b> as the starting
point and scaling to half size with <b>pnmscale</b>, we arrive at
the following smoothed image, with shadow, ready to adorn a
Web page:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadout2.jpg" width=268 height=71 alt="Half scale image with shadow">
</table>
</center>
<h4><a href="how-t.html">How it works with translucent shadows</a></h4>
<h4><a href="./"><b>pnmshadow</b> main page</a></h4>
<p>
<hr>
<p>
<address>
by <a href="/">John Walker</a><br>
August 8th, 1997
</address>
</body>
</html>
<html>
<head>
<title>pnmshadow: How it Works in Translucent Mode</title>
</head>
<body>
<center>
<h1><img src="figures/how_title-t.jpg" width=421 height=159 alt="pnmshadow: How it Works in Translucent Mode"></h1>
</center>
<hr>
<p>
This document describes the process, including PBMplus commands
and the intermediate images they create, by
which <b><a href="./">pnmshadow</a></b>
adds translucent shadows when the <b>-t</b> command line
option is specified. A <a href="how.html">companion document</a>
describes how the default black shadows are generated.
<h3>The Starting Point</h3>
Let's start with the following source image, 536 pixels wide and 141
pixels high. We convert the image from whatever form in which
it was originally created (GIF, JPEG, etc.) to a PPM file before
processing it with <b>pnmshadow</b>.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadin.gif" width=536 height=141 alt="Input image">
</table>
</center>
<h3>The Blank Background</h3>
We start by determining the size of the input image with
<b>pnmfile</b> and then constructing an image with the same
size as the input image consisting entirely of the background
color, which is defined as the color of the pixel at the upper
left corner of the source image. This is performed by the
command:
<p>
<pre>
pnmcut 0 0 1 1 <em>ifile</em> | pnmscale -xsize <em>xsize</em> -ysize <em>ysize</em> ><em>fname</em>-5.ppm
</pre>
<p>
yielding the image:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt5.gif" width=536 height=141 alt="Blank background image">
</table>
</center>
<h3>The Positive Mask</h3>
A positive mask image is created in which all pixels of the background
color are set to white and all other pixels are black. This is accomplished
by subtracting the blank background image from the input (using the
<tt>-difference</tt> option on <b>pnmarith</b> to avoid clipping at
zero or the maximum pixel value), then inverting the result and
thresholding it to a monochrome bitmap.
<p>
<pre>
pnmarith -difference <em>ifile</em> <em>fname</em>-5.ppm | pnminvert | ppmtopgm | pgmtopbm -thresh -value 1.0 ><em>fname</em>-1.ppm
</pre>
<p>
This produces the following mask image.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt1.gif" width=536 height=141 alt="Positive mask image">
</table>
</center>
<h3>The Blurred Image</h3>
Since we wish to simulate a shadow from a nearby extended
light source rather than a sharp shadow as cast by the
Sun, we need to prepare a blurred version of the original
image.
A convolution kernel which averages
the number of pixels specified by the <b>-b</b> option
(default 11), written into the temporary file <tt><em>fname</em>-2.ppm</tt>
in ASCII PGM format, and then the blurred image is created with
the command:
<p>
<pre>
pnmconvol <em>fname</em>-2.ppm <em>ifile</em> ><em>fname</em>-10.ppm
</pre>
<p>
With the default blur setting of 11 pixels, the blurred image
below is generated.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt10-t.jpg" width=536 height=141 alt="Blurred shadow">
</table>
</center>
<h3>Offset Shadow Clip</h3>
Shadows, even bogus ones like we're generating, usually look best when
cast by a light source diagonally displaced from the centre of the
shadow-casting object. To achieve this effect, we first cut a
rectangle from the blurred shadow image reduced in size by the
the number of pixels specified by the <b>-b</b> option
which default to half the blur (<b>-b</b>) setting. The
<em>xsize</em> and <em>ysize</em> arguments in the following
command are the size of the input image in pixels less the shadow
displacement in the respective axis.
<p>
<pre>
pnmcut 0 0 <em>xsize</em> <em>ysize</em> <em>fname</em>-10.ppm ><em>fname</em>-4.ppm
</pre>
<p>
The shadow clip is the identical to the shadow on background color, but
smaller by the offset in each direction.
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt4-t.jpg" width=531 height=136 alt="Offset shadow clip">
</table>
</center>
<h3>Offset Shadow</h3>
Now we're ready to assemble the shadow offset by the specified number
of pixels. We do this by pasting the image cut in the previous step
into the blank background, yielding an image the same size as the
source image with the blurred shadow displaced to the right and
down.
<p>
<pre>
pnmpaste -replace <em>fname</em>-4.ppm <em>xoffset</em> <em>yoffset</em> <em>fname</em>-5.ppm ><em>fname</em>-6.ppm
</pre>
<p>
This gives the following result:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt6-t.jpg" width=536 height=141 alt="Offset shadow">
</table>
</center>
<h3>Inverse Mask</h3>
In order to stitch everything together, we need an inverse of the
mask prepared earlier--one where black pixels represent the background
and all other material is white. This is easily accomplished by
running the positive mask through <b>pnminvert</b>:
<p>
<pre>
pnminvert <em>fname</em>-1.ppm ><em>fname</em>-7.ppm
</pre>
<p>
yielding:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt7.gif" width=536 height=141 alt="Inverse mask">
</table>
</center>
<h3>Masked Input Image</h3>
Now we use the inverse mask prepared in the previous step to create
an image containing all non-background pixels from the source image,
with background pixels set to black. We simply multiply the
inverse mask by the source image:
<p>
<pre>
pnmarith -multiply <em>ifile</em> <em>fname</em>-7.ppm ><em>fname</em>-8.ppm
</pre>
<p>
<em>et voilà:</em>
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt8.gif" width=536 height=141 alt="Masked Input Image">
</table>
</center>
<h3>Shadow with Source Masked</h3>
Our last intermediate step before joining the image with its
shadow is preparing a shadow image with all non-background pixels
in the source image set to black. This ensures that when we add
the image and the shadow, the shadow will not override any pixel
in the source image.
<p>
<pre>
pnmarith -multiply <em>fname</em>-6.ppm <em>fname</em>-1.ppm ><em>fname</em>-9.ppm
</pre>
<p>
This is accomplished by multiplying the shadow by the positive
mask image, which sets all non-background pixels in the source
image to black:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadt9-t.jpg" width=536 height=141 alt="Shadow with Source Masked">
</table>
</center>
<h3>The Final Product</h3>
At long last, we're ready to put together the pieces and deliver
the image to our ever-patient user. This amounts simply to
adding the masked input image (consisting solely of non-background
pixels from the original image) to the shadow with source masked
(in which all source pixels are black):
<p>
<pre>
pnmarith -add <em>fname</em>-8.ppm <em>fname</em>-9.ppm
</pre>
<p>
The resulting image, with shadow, is as follows:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadout-t.jpg" width=536 height=141 alt="Output: Image with translucent shadow">
</table>
</center>
<h3>Smooth Operator</h3>
Since many computer graphics programs create sharp edges on
text, it's often best to create an image at a greater resolution
than that used for presentation, then scale it to the final
resolution with a tool which resamples the image, thus
minimising jagged edges by averaging adjacent
pixels. Using the output of <b>pnmshadow</b> as the starting
point and scaling to half size with <b>pnmscale</b>, we arrive at
the following smoothed image, with shadow, ready to adorn a
Web page:
<p>
<center>
<table border=5>
<tr><td>
<img src="figures/shadout2-t.jpg" width=268 height=71 alt="Half scale image with translucent shadow">
</table>
</center>
<h4><a href="how.html">How it works with black shadows</a></h4>
<h4><a href="./"><b>pnmshadow</b> main page</a></h4>
<p>
<hr>
<p>
<address>
by <a href="/">John Walker</a><br>
August 8th, 1997
</address>
</body>
</html>