filter

filter (or fi) subcommand applies a convolutive filter on a a raster image.

$ rio georastertools filter --help
usage: rio georastertools filter [-h] {median,sum,mean,adaptive_gaussian} ...

Apply a filter to a set of images.

optional arguments:
  -h, --help            show this help message and exit

Filters:
  {median,sum,mean,adaptive_gaussian}
    median              Apply median filter
    sum                 Apply local sum filter
    mean                Apply local mean filter
    adaptive_gaussian   Apply adaptive gaussian filter

The available filters are Adaptive Gaussian, Local Sum, and Local Mean. Each filter is used as a sub-command and has specific arguments for filtering. To see the definitions of these arguments, type the option –help.

• Median

  $ rio georastertools filter median --help
    usage: rio georastertools filter median [-h] --kernel_size KERNEL_SIZE [-o OUTPUT]
                                     [-ws WINDOW_SIZE]
                                     [-p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}]
                                     [-b BANDS [BANDS ...]] [-a]
                                     inputs [inputs ...]
  
    Apply a median filter (see scipy median_filter for more information)
  
    positional arguments:
      inputs                Input file to process (e.g. Sentinel2 L2A MAJA from
                            THEIA). You can provide a single file with extension
                            ".lst" (e.g. "filtering.lst") that lists the input
                            files to process (one input file per line in .lst)
  
    optional arguments:
      -h, --help            show this help message and exit
      --kernel_size KERNEL_SIZE
                            Kernel size of the filter function, e.g. 3 means a
                            square of 3x3 pixels on which the filter function is
                            computed (default: 8)
      -o OUTPUT, --output OUTPUT
                            Output dir where to store results (by default current
                            dir)
      -ws WINDOW_SIZE, --window_size WINDOW_SIZE
                            Size of tiles to distribute processing, default: 1024
      -p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}, --pad {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}
                            Pad to use around the image, default : edge (see https
                            ://numpy.org/doc/stable/reference/generated/numpy.pad.
                            html for more information)
      -b BANDS [BANDS ...], --bands BANDS [BANDS ...]
                            List of bands to compute
      -a, --all             Compute all bands
  
    By default only first band is computed.
  

  The corresponding API functions that is called by the command line interface is the following :

  eolab.georastertools.processing.algo.median(input_data: ndarray, kernel_size: int) → ndarray

  Applies a Median Filter to the input data using scipy.ndimage.median_filter. The filter computes the median of the values contained within a sliding window determined by the kernel size.

  Parameters:

  • input_data (np.ndarray) – A 3D numpy array of shape (1, number_of_lines, number_of_columns) containing the Digital Elevation Model (DEM). The function only accepts arrays with one band.

  • kernel_size (int) – The size of the sliding window (kernel) used to compute the median.

  Returns:

  A numpy array of the same shape as input_data, containing the filtered data with the median values computed in the specified kernel.

  Return type:

  np.ndarray

  Raises:

  ValueError – If input_data does not have 3 dimensions or if the first dimension is not of size 1, or if kernel_size is not a positive odd integer.

  Here is an example of a median filter applied to the NDVI of a SENTINEL2 L2A THEIA image cropped to a region of interest. This raster was previously computed using radioindice on the original SENTINEL2 L2A THEIA image.

  $ rio georastertools filter median --kernel_size 16 "./SENTINEL2A_20180928-105515-685_L2A_T30TYP_D-ndvi.tif"
  

  Original	Filtered by Median

• Local sum

  $ rio georastertools filter sum --help
    usage: rio georastertools filter sum [-h] --kernel_size KERNEL_SIZE [-o OUTPUT]
                                  [-ws WINDOW_SIZE]
                                  [-p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}]
                                  [-b BANDS [BANDS ...]] [-a]
                                  inputs [inputs ...]
  
    Apply a local sum filter using integral image method
  
    positional arguments:
      inputs                Input file to process (e.g. Sentinel2 L2A MAJA from
                            THEIA). You can provide a single file with extension
                            ".lst" (e.g. "filtering.lst") that lists the input
                            files to process (one input file per line in .lst)
  
    optional arguments:
      -h, --help            show this help message and exit
      --kernel_size KERNEL_SIZE
                            Kernel size of the filter function, e.g. 3 means a
                            square of 3x3 pixels on which the filter function is
                            computed (default: 8)
      -o OUTPUT, --output OUTPUT
                            Output dir where to store results (by default current
                            dir)
      -ws WINDOW_SIZE, --window_size WINDOW_SIZE
                            Size of tiles to distribute processing, default: 1024
      -p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}, --pad {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}
                            Pad to use around the image, default : edge (see https
                            ://numpy.org/doc/stable/reference/generated/numpy.pad.
                            html for more information)
      -b BANDS [BANDS ...], --bands BANDS [BANDS ...]
                            List of bands to compute
      -a, --all             Compute all bands
  
    By default only first band is computed.
  

  The corresponding API functions that is called by the command line interface is the following :

  eolab.georastertools.processing.algo.local_sum(input_data: ndarray, kernel_size: int = 8) → ndarray

  Computes the local sums of the input data using a sliding window defined by the kernel size. Each element in the output is the sum of the pixels within the specified kernel size window.

  Parameters:

  • input_data (np.ndarray) – A 3D numpy array of shape (1, number_of_lines, number_of_columns) containing the Digital Elevation Model (DEM). The function only accepts arrays with one band.

  • kernel_size (int) – The size of the sliding window used to compute the local sum.

  Returns:

  A numpy array of the same size as input_data containing the computed local sums.

  Return type:

  np.ndarray

  Raises:

  ValueError – If input_data does not have 3 dimensions or if the first dimension is not of size 1.

  Here is an example of the local mean applied to the NDVI of a SENTINEL2 L2A THEIA image cropped to a region of interest. This raster was previously computed using radioindice on the original SENTINEL2 L2A THEIA image.

  $ rio georastertools filter sum --kernel_size 16 "./SENTINEL2A_20180928-105515-685_L2A_T30TYP_D-ndvi.tif"
  

  Original	Filtered by Local sum

• Local mean

  $ rio georastertools filter mean --help
    usage: rio georastertools filter mean [-h] --kernel_size KERNEL_SIZE [-o OUTPUT]
                           [-ws WINDOW_SIZE]
                           [-p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}]
                           [-b BANDS [BANDS ...]] [-a]
                           inputs [inputs ...]
  
    Apply a local mean filter using integral image method
  
    positional arguments:
      inputs                Input file to process (e.g. Sentinel2 L2A MAJA from
                            THEIA). You can provide a single file with extension
                            ".lst" (e.g. "filtering.lst") that lists the input
                            files to process (one input file per line in .lst)
  
    optional arguments:
      -h, --help            show this help message and exit
      --kernel_size KERNEL_SIZE
                            Kernel size of the filter function, e.g. 3 means a
                            square of 3x3 pixels on which the filter function is
                            computed (default: 8)
      -o OUTPUT, --output OUTPUT
                            Output dir where to store results (by default current
                            dir)
      -ws WINDOW_SIZE, --window_size WINDOW_SIZE
                            Size of tiles to distribute processing, default: 1024
      -p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}, --pad {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}
                            Pad to use around the image, default : edge (see https
                            ://numpy.org/doc/stable/reference/generated/numpy.pad.
                            html for more information)
      -b BANDS [BANDS ...], --bands BANDS [BANDS ...]
                            List of bands to compute
      -a, --all             Compute all bands
  
    By default only first band is computed.
  

  The corresponding API functions that is called by the command line interface is the following :

  eolab.georastertools.processing.algo.local_mean(input_data: ndarray, kernel_size: int = 8) → ndarray

  Computes the local means of the input data using a sliding window defined by the kernel size. Each element in the output is the mean of the pixels within the specified kernel size window.

  Parameters:

  • input_data (np.ndarray) – A 3D numpy array of shape (1, number_of_lines, number_of_columns) containing the Digital Elevation Model (DEM). The function only accepts arrays with one band.

  • kernel_size (int) – The size of the sliding window used to compute the local mean.

  Returns:

  A numpy array of the same size as input_data containing the computed local means.

  Return type:

  np.ndarray

  Raises:

  ValueError – If input_data does not have 3 dimensions or if the first dimension is not of size 1.

  Here is an example of the local mean applied to the NDVI of a SENTINEL2 L2A THEIA image cropped to a region of interest. This raster was previously computed using radioindice on the original SENTINEL2 L2A THEIA image.

  $ rio georastertools filter mean --kernel_size 16 "./SENTINEL2A_20180928-105515-685_L2A_T30TYP_D-ndvi.tif"
  

  Original	Filtered by Local mean

• Adaptative gaussian

  $ rio georastertools filter adaptive_gaussian --help
  usage: rio georastertools filter adaptive_gaussian [-h] --kernel_size KERNEL_SIZE
                                              --sigma SIGMA [-o OUTPUT]
                                              [-ws WINDOW_SIZE]
                                              [-p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}]
                                              [-b BANDS [BANDS ...]] [-a]
                                              inputs [inputs ...]
  
  Apply an adaptive (Local gaussian of 3x3) recursive filter on the input image
  
  positional arguments:
    inputs                Input file to process (e.g. Sentinel2 L2A MAJA from
                          THEIA). You can provide a single file with extension
                          ".lst" (e.g. "filtering.lst") that lists the input
                          files to process (one input file per line in .lst)
  
  optional arguments:
    -h, --help            show this help message and exit
    --kernel_size KERNEL_SIZE
                          Kernel size of the filter function, e.g. 3 means a
                          square of 3x3 pixels on which the filter function is
                          computed (default: 8)
    --sigma SIGMA         Standard deviation of the Gaussian distribution
                          (sigma)
    -o OUTPUT, --output OUTPUT
                          Output dir where to store results (by default current
                          dir)
    -ws WINDOW_SIZE, --window_size WINDOW_SIZE
                          Size of tiles to distribute processing, default: 1024
    -p {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}, --pad {none,edge,maximum,mean,median,minimum,reflect,symmetric,wrap}
                          Pad to use around the image, default : edge (see https
                          ://numpy.org/doc/stable/reference/generated/numpy.pad.
                          html for more information)
    -b BANDS [BANDS ...], --bands BANDS [BANDS ...]
                          List of bands to compute
    -a, --all             Compute all bands
  
  By default only first band is computed.
  

  The corresponding API functions that is called by the command line interface is the following :

  eolab.georastertools.processing.algo.adaptive_gaussian(input_data: ndarray, kernel_size: int = 8, sigma: int = 1) → ndarray

  Applies an Adaptive Gaussian Filter to the input data that smoothes the input while preserving edges.

  Parameters:

  • input_data (np.ndarray) – A 3D numpy array of shape (1, number_of_lines, number_of_columns) containing the Digital Elevation Model (DEM). The function only accepts arrays with one band.

  • kernel_size (int) – The size of the kernel used for the adaptive filtering. Default is 8.

  • sigma (int) – The standard deviation of the Gaussian distribution, which controls the level of smoothing. Default is 1.

  Returns:

  A numpy array of the same shape as input_data, containing the filtered data.

  Return type:

  np.ndarray

  Raises:

  ValueError – If input_data does not have 3 dimensions or if the first dimension is not of size 1.

  Here is an example of the local mean applied to the NDVI of a SENTINEL2 L2A THEIA image cropped to a region of interest. This raster was previously computed using radioindice on the original SENTINEL2 L2A THEIA image.

  $ rio georastertools filter adaptive_gaussian --kernel_size 16 --sigma 1 "./SENTINEL2A_20180928-105515-685_L2A_T30TYP_D-ndvi.tif"
  

  Original	Filtered by Adaptive gaussian