À̸§°Ë»ö
¸ÞÀϺ¸³»±â

À̸§°Ë»ö
À̸§°Ë»ö
À̸§°Ë»ö
À̸§°Ë»ö
¸ÞÀϺ¸³»±â

À̸§°Ë»ö
¸ÞÀϺ¸³»±â

À̸§°Ë»ö
À̸§°Ë»ö
¸ÞÀϺ¸³»±â

À̸§°Ë»ö
À̸§°Ë»ö
¸ÞÀϺ¸³»±â

À̸§°Ë»ö

::: eSAR Forum :::


20 12 Åë°èÄ«¿îÅÍ º¸±â   ȸ¿ø °¡ÀÔ È¸¿ø ·Î±×ÀÎ °ü¸®ÀÚ Á¢¼Ó --+
Name   Hoonyol
Subject   ¢Ñ question about the range_migration.c
¦­[minghui hou said:]
¦­I have some questions about your esar package_ range-migration.c.
¦­1. I do not understand the formulation in line 44 :az_ambiguity = floor(f_Dc_middle/prf+0.5);
¦­How did you derive it?

-> Doppler_centroid = m * PRF + D_c
where the absolute value of D_c is smaller than PRF/2 and m is the azimuth ambiguity. For more detailed infomation about azimuth ambiguity please refer to SAR textbook such as [Curlander and McDonough, 1991]

¦­2. your interpolation uses the nearest point method, dont you? when does it need to use sinc inperpolation?

-> We always need sinc interpolation which is not very difficult to implement. This would be in the first line of the upgrade plan.

¦­3.in Line 116, can you give me some tips about why do such alignment? I have not got enough
¦­backgroud knowledge about why do so.
¦­

-> This alignment is necessary for azimuth matched filtering, for the exact match with azimuth chirp signal. You can skip this alignment here but you need to do it before azimuth matched filtering anyway. There can also be some alternatives keeping frequency and time domain matched filtering correctly. I agree this alignment need not present in the range-migration.c and might deserve a separate file or can be merged into azimuth matched filtering.


¦­by the way, I am studying sar, but have not understand it well, so please dont lose patience
¦­for me, I will catch up as soon as possible.
¦­
¦­thank you for your time and kindness

-> You are very welcome.

¦­
¦­I noticed that the output from your fft is different that of matlab as following:
¦­
¦­the matlab fft function output order is : 1 2    3    ... 1023 1024;
¦­your fft output order is :                1 1024 1023 ...         2;
¦­

-> This is related to the definition of Fourier Transform and FFT algorithm I have used. As mentioned in the program, I used the one from Numerical Recipes in C. It uses positive sense for FFT i.e. exp(j*2*pi*f*t)  instead of exp(-j*2*pi*f*t). The result is flipping function in time or frequency domain. As eSAR use iFFT followed by FFT, it doesn't cause any problem in producing the final result. One thing you need be careful is during the range migration where + or - should be cautiously chosen for migration.

Anyway, could you please double check that my FFT output order is [1024 1023 ... 2 1]? Otherwise some errors have crept in my source code and it would have been a nice decision to open this source code because you helped me in return.

¦­How will this differenc affect your choise of chirp data structure in the range_chirp.c?
¦­
¦­<< Data Structure>>
¦­
¦­        *****...****0   ... 0 ...     0*****...****
¦­        -------------------------------------------
¦­fast-        |          |           |          |  
¦­ time(t):   0    chirp_duration/2  -chirp_duration/2 -delta_t        
¦­
¦­range_chirp |          |                   |          |
¦­_data[i]:   0      chirp_end        chirp_start  nrange_fft-1
¦­
¦­Why the zeros are padded in the middle of the array, not its end?


-> This is how FFT works... well, as far as I understand. FFT assumes that All input functions are periodic. That's how you put data of negative time, isn't it?

Cheers,
Hoonyol

°Ô½Ã¹°À» À̸ÞÀÏ·Î º¸³»±â ÇÁ¸°Æ®Ãâ·ÂÀ» À§ÇÑ È­¸éº¸±â
DATE: 2005.08.19 - 21:26

210.115.53.203 - Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)


 ÀÌÀü±Û question about the range_migration.c
 ´ÙÀ½±Û ¢Ñ ¢Ñ question about the range_migration.c
±Û³²±â±â»èÁ¦Çϱâ¼öÁ¤Çϱâ´äº¯´Þ±âÀüü ¸ñ·Ï º¸±â

üũµÈ Ç׸ñ ÇѲ¨¹ø¿¡ º¸±â
20Simple vieweSAR¿¡ »ç¿ëµÈ  _fftº¯¼ö °ü·Ã Q&A [1] ±è¼ö¹Ì 2006.09.12 24304 *
19Simple viewSAR °³³ä¿¡ °ü·ÃµÈ ¸î°¡Áö Áú¹® [1] shs 2006.04.22 25363 *
18Simple viewDisplay result data from eSAR Lee Chol 2006.04.04 23356 *
17Simple view¢Ñ Display result data from eSAR ÀÌÈÆ¿­ 2006.04.04 25517 *
16Simple viewthe discussion about raw data structure of rada... minghui ho... 2005.09.01 25046 *
15Simple viewquestion about the range_migration.c minghui ho... 2005.08.19 29656 *
14ÇöÀç Àаí ÀÖ´Â ±ÛÀÔ´Ï´Ù.¢Ñ question about the range_migration.c Hoonyol 2005.08.19 23119 *
13Simple view¢Ñ ¢Ñ question about the range_migration.c minghui ho... 2005.08.20 28064 *
12Simple view¢Ñ ¢Ñ ¢Ñ question about the range_migration.c Hoonyol 2005.08.22 28869 *
11Simple viewnominal mean and dc bias nauman 2005.08.15 28240 *
üũµÈ Ç׸ñ ÇѲ¨¹ø¿¡ »èÁ¦/º¹»ç/À̵¿ Çϱâ
üũµÈ Ç׸ñ »èÁ¦ üũµÈ Ç׸ñ »èÁ¦
üũµÈ Ç׸ñ À̵¿ üũµÈ Ç׸ñ À̵¿
üũµÈ Ç׸ñ º¹»ç üũµÈ Ç׸ñ º¹»ç
ÇöÀçÆäÀÌÁö°¡ ùÆäÀÌÁö ÀÔ´Ï´Ù. ´ÙÀ½ÆäÀÌÁö
ÀÌÀü 1  2 ´ÙÀ½
±Û³²±â±â »õ·Î°íħ
À̸§À» °Ë»öÇ׸ñ¿¡ Ãß°¡/Á¦°ÅÁ¦¸ñÀ» °Ë»öÇ׸ñ¿¡ Ãß°¡/Á¦°Å³»¿ëÀ» °Ë»öÇ׸ñ¿¡ Ãß°¡/Á¦°Å ¸ÞÀÎÈ­¸éÀ¸·Î µ¹¾Æ°¡±â