ERDAS OrthoBASE 正射纠正SPOT5影像

<P  align=justify><FONT face=Arial>ERDAS IMAGINE OrthoBASE8.4</FONT>和<FONT face=Arial>ERDAS IMAGINE OrthoBASE 8.5</FONT>中提供的<FONT face=Arial>Generic Pushbroom</FONT>传感器模型可以精确地纠正<FONT face=Arial>SPOT5</FONT>的<FONT face=Arial>10</FONT>米、<FONT face=Arial>5</FONT>米、<FONT face=Arial>2.5</FONT>米的全色和多光谱影像。</P>
<P  align=justify>纠正时需要输入的关键参数如下，见<FONT face=Arial>1</FONT>、<FONT face=Arial>2</FONT>。纠正精度见<FONT face=Arial>4</FONT>。</P>
<P  align=justify><b><FONT face=Arial>1. Sensor Information:</FONT></b></P>
<TABLE height=126 cellSpacing=3 cellPadding=0 width=500 border=0>

<TR>
<TD width=209 height=23>
<P align=right><FONT face=Arial>Focal Length: </FONT></P></TD>
<TD width=291 height=23><FONT face=Arial>1082 mm</FONT></TD></TR>
<TR>
<TD vAlign=top width=209 height=23><FONT face=Arial>
<P align=right>Pixel Size:</FONT></P></TD>
<TD width=291 height=23><FONT face=Arial>0.013 mm (10 meter MSS)
0.0065 mm (5 meter Pan)
0.00325 mm (2.5 meter Pan)</FONT></TD></TR>
<TR>
<TD vAlign=top width=209 height=23><FONT face=Arial>
<P align=right>Sensor columns (full scene, 1A): </FONT></P></TD>
<TD width=291 height=23><FONT face=Arial>available from the metadata
6000 (10 meter MSS)
12000 (5 meter Pan)
24000 (2.5 meter Pan)</FONT></TD></TR></TABLE>
<P align=justify><b><FONT face=Arial>2. Frame Attributes:</FONT></b></P>
<TABLE cellSpacing=5 cellPadding=0 width=500 border=0>

<TR>
<TD vAlign=top width=188>
<P align=right><FONT face=Arial>Side Incidence (degrees):</FONT></P></TD>
<TD width=300><FONT face=Arial>available from the metadata
under title INCIDENCE_ANGLE
KEY: negate whatever value there!!!
</FONT>注意：在元数据文件中，如果传感器的倾角（<FONT face=Arial>Incidence-angle)</FONT>为“<FONT face=Arial>R8.006</FONT>”，侧在上面的参数设置中<b><FONT color=#ff0000><FONT face=Arial>Side incidence=-8.006</FONT>，即为负值</FONT></b>。</TD></TR>
<TR>
<TD vAlign=top width=188>
<P align=right><FONT face=Arial>Ground Resolution: </FONT></P></TD>
<TD width=300><FONT face=Arial>available from the metadata
10 m (10 meter MSS)
5 m (5 meter Pan)
2.5 m (2.5 meter Pan)</FONT></TD></TR></TABLE>
<P align=justify><b><FONT face=Arial>3. Accurate and evenly distributed points (about 12)</FONT></b></P>
<P align=justify><b><FONT face=Arial>4. Here are the RMSEs I got from one SPOT 5 data set (stereo pair) in our archive:</FONT></b></P>
<TABLE height=93 cellSpacing=5 cellPadding=0 width=500 border=0>

<TR>
<TD align=middle width=107 height=15></TD>
<TD align=middle width=173 height=15><FONT face=Arial>10 Ground Controls</FONT></TD>
<TD align=middle width=202 height=15><FONT face=Arial>6 Check Points</FONT></TD></TR>
<TR>
<TD align=middle width=107 height=18>
<P align=center><FONT face=Arial>x</FONT></P></TD>
<TD align=middle width=173 height=18><FONT face=Arial>0.11 pixels</FONT></TD>
<TD align=middle width=202 height=18><FONT face=Arial>0.18 pixels</FONT></TD></TR>
<TR>
<TD align=middle width=107 height=18><FONT face=Arial>y</FONT></TD>
<TD align=middle width=173 height=18><FONT face=Arial>0.79 pixels</FONT></TD>
<TD align=middle width=202 height=18><FONT face=Arial>0.8 pixels</FONT></TD></TR>
<TR>
<TD align=middle width=107 height=18><FONT face=Arial>z</FONT></TD>
<TD align=middle width=173 height=18><FONT face=Arial>1.7 meters</FONT></TD>
<TD align=middle width=202 height=18><FONT face=Arial>3.9 meters</FONT></TD></TR></TABLE>

这个有点老了，这种方法在高差较大的地方误差还是很大的。但ERDAS87已经可以较好的支持SPOT5。

<img src="images/post/smile/dvbbs/em02.gif" /><img src="images/post/smile/dvbbs/em01.gif" />

<img src="images/post/smile/dvbbs/em01.gif" /><img src="images/post/smile/dvbbs/em01.gif" />

ok

<img src="images/post/smile/dvbbs/em01.gif" />

<img src="images/post/smile/dvbbs/em04.gif" />

<img src="images/post/smile/dvbbs/em01.gif" /><img src="images/post/smile/dvbbs/em02.gif" />