Skip to content

a-ngo/radar-signal-processing

Repository files navigation

Project

Visualization

1. FFT 2. FFT 2D CFAR

2D CFAR

  1. Implementation steps for the 2D CFAR process
  • loop over all bins (range and doppler)
  • slice the range doppler map to create the training cells
  • and convert them to power
  • calculate mean noise level of the training cells
  • revert cells back to decibel
  • add an offset to keep the false alarm to the minimum
  • finally apply threshold
  1. Selection of Training, Guard cells and offset
  • the parameters are defined starting from line 146-157
  • I chose them by trial and error
  • choosing the right offset was essential for a sufficient result
  1. Steps taken to suppress the non-thresholded cells at the edges
  • see line 187
  • I just setted every cell to zero
  • and only add a value to the CFAR result if the threshold is exceeded

Releases

No releases published

Packages

No packages published