Step 0.将fits转为bdf

Midas 050> INDISK/FITS pdc081211_0021.fits pdc081211_0021_B.bdf

执行nametrans.prg可以快速将所有fits转bdf

Midas 051> @@ nametrans.prg

Step I. 逐幅检查图像看坏的column.

Midas 015> CREATE/DISP

Midas 015> LOAD/IMAG   pdc081211_0021_B.bdf

Note:有时导入图像可能感觉什么也没有，在后面加上cuts=f改善；

Midas 102> LOAD/IMAG   pdc081211_0028_B.bdf cuts=f

Midas 015> CREATE/GRAP

Midas 021> PLOT/COLU   pdc081211_0022_B.bdf

调整Pixel Value的范围,显示效果明显变好，看到图中的Bad pixel.

Midas 025> LOAD/IMAG   pdc081211_0022_B.bdf cuts=3970,4000

Midas 028> EXTRAC/CTRA ? ? plot            

Min, Max = 3984, 4165    at x-pixel #570  !中间亮线

Midas 028> EXTRAC/CTRA ? ? plot            

Min, Max = 3983, 4033    at x-pixel #737  !右边稍亮的线

Midas 028> EXTRAC/CTRA ? ? plot            

Min, Max = 3984, 4194    at x-pixel #408  !左边的亮点

Step II. 合并所有图像看得更清楚

1.创建要合并图像的输入incatalog.cat  

Midas 317> CREATE/ICAT incatalog pdc081211_0*_B.bdf

Midas 317> SET/ICAT incatalog.cat

Midas 317> AVERAG/IMAG BIASav=incatalog.cat

Midas 317> LOAD/IMAG   BIASav.bdf cuts=f

合并的图像中可以明显看到有3条坏的column

2.找出这些坏column的坐标

Midas 329> EXTRAC/CTRA ? ? plot

点选坏的列，窗口会显示列坐标:

x-pixel #466   !别外找了3个:#5706  #7376  #1201

另一方法(推荐):

Midas 329> GET/CURS    

frame: BIASav.bdf  (data = R4)                                                

      plane_no 1 loaded

cursor #0

    frame pixels           world coords        intensity

    467      279       467.000     279.000       3789.65  

    466      340       466.000     340.000       3790.15  

    465      368       465.000     368.000       3789.40  

    570      224       570.000     224.000       3802.30  

    570      246       570.000     246.000       3801.10  

    737      214       737.000     214.000       3791.95  

   1201      215       1201.00     215.000       3790.65  

   1201      254       1201.00     254.000       3791.35  

确定有3个:#466  #570  #737  #1201  第一个不太时显

Step III. 改正图像的这三列

1.一个column改正:

Midas 017> COMPUT/COLU BIASav.C466 = ((1*C465)+(1*C467))/2.0

Midas 019> COMPUT/COLU BIASav.C570 = ((1*C569)+(1*C571))/2.0

Midas 028> COMPUT/COLU BIASav.C737 = ((1*C736)+(1*C738))/2.0

Midas 031> COMPUT/COLU BIASav.C1201 = ((1*C1200)+(1*C1202))/2.0

2.三个column改正:

Midas 023> COMPUT/COLU BIASav.C569 = ((3*C568)+(1*C572))/4.0

Midas 024> COMPUT/COLU BIASav.C570 = ((2*C568)+(2*C572))/4.0

Midas 025> COMPUT/COLU BIASav.C571 = ((1*C568)+(3*C572))/4.0

3.五个column改正:

Midas 023> COMPUT/COLU BIASav.C568 = ((5*C567)+(1*C573))/6.0

Midas 023> COMPUT/COLU BIASav.C569 = ((4*C567)+(2*C573))/6.0

Midas 024> COMPUT/COLU BIASav.C570 = ((3*C567)+(3*C573))/6.0

Midas 025> COMPUT/COLU BIASav.C571 = ((2*C567)+(4*C573))/6.0

Midas 025> COMPUT/COLU BIASav.C572 = ((1*C567)+(5*C573))/6.0

Midas 032> LOAD/IMAG   BIASav.bdf scale=1,1 cuts=f

可以看到改正之后的图像.

Step IV:改正每一幅图像的bad column

建一个prg文件: c_bad_col.prg

COMPUT/COLU pdc081211_0021_B.C570 = ((1*C569)+(1*C571))/2.0

COMPUT/COLU pdc081211_0022_B.C570 = ((1*C569)+(1*C571))/2.0

COMPUT/COLU pdc081211_0023_B.C570 = ((1*C569)+(1*C571))/2.0

COMPUT/COLU pdc081211_0024_B.C570 = ((1*C569)+(1*C571))/2.0

...

执行改正:

Midas 041> @@ c_bad_col.prg

References:

1. http://www.eso.org/sci/software/esomidas//doc/user/98NOV/vola/node19.html

                     2015-10-26 17:45 ROB Brussels,Belgium