forked from markummitchell/engauge-digitizer
-
Notifications
You must be signed in to change notification settings - Fork 0
/
answerdiscretizing.html
78 lines (78 loc) · 3.52 KB
/
answerdiscretizing.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<HTML>
<HEAD>
<TITLE>Engauge Digitizer - Discretizing</TITLE>
<META NAME="keywords" CONTENT="Engauge Digitizer, browser based documentation, online manual, Discretizing">
<META NAME="description" CONTENT="Discretizing a graph or map">
</HEAD>
<BODY>
<H2>
<A HREF="index.html">
<IMG SRC="animation.png" ALIGN=MIDDLE ALT="" STYLE="border: 0px"></A>
Engauge Digitizer - Discretizing</H2>
<HR WIDTH="100%">
<H2>What Is Discretizing?</H2>
<P>When an image is imported, Engauge Digitizer processes that image
by converting each pixel into black or white. Since most images use
black writing on a white background, the black pixels are
assumed to represent curves, and the white pixels are assumed to
represent the background. This process of converting color images into
black foreground and white background is called <em>discretizing</em>.
The resulting processed image can be automatically digitized
using <A HREF="tutorautolinegraph.html#segmentfill">Segment Fill</A> and
<A HREF="tutorautopointgraph.html#pointmatch">Point Match</A>.</P>
<P>Usually the default discretizing settings correctly differentiate
the foreground from the background. For example, in the two images
below, the original image on the left gets processed into the image
on the right.</P>
<TABLE ALIGN="CENTER">
<TR>
<TD><IMG SRC="discretize_ok_before.png" ALT="OK before" WIDTH="247" HEIGHT="208"></TD>
<TD><IMG SRC="discretize_ok_after.png" ALT="OK after" WIDTH="247" HEIGHT="208"></TD>
</TR>
<TR>
<TD ALIGN="CENTER">Before discretizing</TD>
<TD ALIGN="CENTER">After discretizing</TD>
</TR>
</TABLE>
<P>However, an image that uses color might not get correctly digitized. In
the two images below, the original image on the left is processed into
the featureless image on the right. The relatively dark green was
assumed to be part of the foreground.</P>
<TABLE ALIGN="CENTER">
<TR>
<TD><IMG SRC="discretize_bad_color_before.png" ALT="OK before" WIDTH="247" HEIGHT="208"></TD>
<TD><IMG SRC="discretize_bad_color_after.png" ALT="OK after" WIDTH="247" HEIGHT="208"></TD>
</TR>
<TR>
<TD ALIGN="CENTER">Before discretizing</TD>
<TD ALIGN="CENTER">After discretizing</TD>
</TR>
</TABLE>
<P>Even a black-and-white image might be discretized in a manner that
is not very helpful. In the two images below, the original image on
the left is processed into the image on the right.
One curve is so thin it has broken up into many unconnected
short segments, and the other curve is so light that it has
disappeared entirely! Neither curve will be recognized
by the curve recognition algorithms in Engauge Digitizer.</P>
<TABLE ALIGN="CENTER">
<TR>
<TD><IMG SRC="discretize_bad_gray_before.png" ALT="OK before" WIDTH="247" HEIGHT="208"></TD>
<TD><IMG SRC="discretize_bad_gray_after.png" ALT="OK after" WIDTH="247" HEIGHT="208"></TD>
</TR>
<TR>
<TD ALIGN="CENTER">Before discretizing</TD>
<TD ALIGN="CENTER">After discretizing</TD>
</TR>
</TABLE>
<H2>How Do I Control Discretizing?</H2>
<P>By simply changing the Discretizing settings, curves can often be extracted
from color, low-contrast and low-resolution images. Such images are often
produced by fax machines, microfiche, cheap scanners, and excessive enlarging
on the copier.</P>
<P>Use the Settings/Discretizing dialog to experiment with various
discretizing settings. You can get more information about the
<A HREF="dlgdiscretize.html">Discretizing dialog</A>.</P>
</BODY>
</HTML>