1

I work mainly in GIS (geographical information systems) but I would like to deepen my knowledge of remote sensing (analyzing imagery acquired remotely from airborne or satellite sensors). I was advised to learn as much about signal analysis as possible. I am assuming that signal processing and analysis refer to the same techniques.

Because signal processing seems to be applied to all sorts of different areas, I'm not quite sure where to start. I should specify that beyond two university calculus courses (on which did quite well), I do not have a background in physics or math.

Could anyone help me plan a sort of "road map" to learning signal processing, preferably applied to remote sensing? Thank you.

AndrewP
  • 11
  • 1
  • A review of integral calculus, complex numbers and trigonometry might help when reading some of the introductory DSP literature. – hotpaw2 Apr 10 '14 at 19:13
  • Classification Methods for Remotely Sensed Data, Second Edition – Emre Oct 07 '14 at 20:46
  • You might also want to have a look at the following responses on this website: http://dsp.stackexchange.com/questions/22264/book-recommendations-on-dsp-statistical-sp-and-signal-detection/22401#22401

    http://dsp.stackexchange.com/questions/427/what-resources-are-recommended-for-an-introduction-to-dsp

    The techniques are broad enough to be applied to different types of signals obtained from different types of sensors (e.g. reflectance spectra)

     – A_A Apr 07 '15 at 11:59

2 Answers2

1

@AndrewP. You have a long road ahead you. But the road is well lit and the rewards will are greater and greater as you travel along. You need to learn digital signal processing (DSP). May I humbly suggest you buy the two books "The Scientist & Engineer's Guide to Digital Signal Processing"

http://www.amazon.com/Scientist-Engineers-Digital-Signal-Processing/dp/0966017633/ref=sr_1_2?ie=UTF8&qid=1438652461&sr=8-2&keywords=scientist%27s+guide+signal+smith

and "Understanding Digital Signal Processing."

http://www.amazon.com/Understanding-Digital-Signal-Processing-3rd/dp/0137027419/ref=ntt_at_ep_dpi_1

They're widely considered to be the best ways to learn DSP for beginners. Do NOT sign up for any free Internet online DSP courses! They will only discourage you at this point in your learning. You'll need to obtain either (1) a 'Student' version of Matlab software or (2) Octave software so you can experiment with DSP on your computer. Do not take any math classes at this point--you can learn selected mathematical topics as necessary as you travel your long road. Andrew if you have enough energy and enthusiasm, to quote Susan B. Anthony, "Failure is impossible."

Richard Lyons
  • 5,777
  • 13
  • 25
  • As an alternative for the Matlab or Octave software you could also have a look into Python and the numpy and scipy modules. – Brian Aug 04 '15 at 06:56
0

My brief recommendation would be to watch the Gilbert Strang Linear Algebra lectures (free MIT courseware, and he also provides a free copy of his textbook in pdf form) to get a comprehensive understanding of linear algebra. Then I would suggest looking into a course (on iTunesU or anywhere else) on Fourier Analysis. That will provide you up with a lot of the math/physics background to understanding signals processing.

Since ArcGIS uses python, you can call any matlab signals processing functions in there if you're looking to use this directly in a GIS.

user3433572
  • 21
  • 4