shareloc.geomodels.grid

Localisation functions from multi h direct grids.

Module Contents

Classes

Grid

multi H direct localization grid handling class.

Functions

coloc(multi_h_grid_src, multi_h_grid_dst, dtm, origin, ...)

colocalization grid on dtm
class shareloc.geomodels.grid.Grid(geomodel_path)[source]

Bases: shareloc.geomodels.geomodel_template.GeoModelTemplate

multi H direct localization grid handling class. please refer to the main documentation grid format

Derives from GeoModelTemplate

Parameters:

  • row0 (float) – grid first pixel center along Y axis (row).

  • col0 (float) – grid first pixel center along X axis (column).

  • nbrow (int) – grid size in row

  • nbcol (int) – grid size in col

  • steprow (float) – grid step in row

  • stepcol (float) – grid step in col

  • rowmax (float) – last row in grid

  • colmax (float) – last col in grid

  • repter (str) – ground coordinate system

  • epsg (int) – epsg code corresponding to shareloc.grid.Grid.repter

  • nbalt (int) – number of altitude layers

  • lon_data (np.ndarray of size (nbalt,nbrow,nbcol) size) – longitude array

  • lat_data (np.ndarray of size (nbalt,nbrow,nbcol) size) – latitude array

  • alts_down (list) – altitudes in decreasing order

  • type (str) – geometric model type

  • geomodel_path (str) –

classmethod load(geomodel_path)[source]

Load grid and fill Class attributes.

The grid is read as an shareloc.image. Image and class attributes are filled. Shareloc geotiff grids are stored by increasing altitude H0 … Hx 2 data cubes are defined: - lon_data : [alt,row,col] - lat_data : [alt,row,col]

read()[source]

Load grid and fill Class attributes.

The grid is read as an shareloc.image. Image and class attributes are filled. Shareloc geotiff grids are stored by increasing altitude H0 … Hx 2 data cubes are defined: - lon_data : [alt,row,col] - lat_data : [alt,row,col]

parse_metadata_alti(metadata)[source]

parse metadata to sort altitude in decreasing order

Parameters:

metadata (dict) – Geotiff metadata

get_alt_min_max()[source]

returns altitudes min and max layers

Returns:

alt_min,lat_max

Return type:

list

direct_loc_h(row, col, alt, fill_nan=False)[source]

direct localization at constant altitude

Parameters:

  • row (float or 1D numpy.ndarray dtype=float64) – line sensor position

  • col (float or 1D numpy.ndarray dtype=float64) – column sensor position

  • alt (float) – altitude

  • fill_nan (boolean) – fill numpy.nan values with lon and lat offset if true (same as OTB/OSSIM), nan is returned otherwise

Returns:

ground position (lon,lat,h)

Return type:

numpy.ndarray 2D dimension with (N,3) shape, where N is number of input coordinates

compute_los(row, col, epsg)[source]

Compute Line of Sight

Parameters:

  • row (float) – line sensor position

  • col (float) – column sensor position

  • epsg (int) – epsg code

Returns:

los

Return type:

numpy.array

direct_loc_dtm(row, col, dtm)[source]

direct localization on dtm

TODO explain algorithm TODO optimize code (for loop, …)

Parameters:

  • row (float) – line sensor position

  • col (float) – column sensor position

  • dtm (shareloc.dtm) – dtm model

Returns:

ground position (lon,lat,h) in dtm coordinates system.

Return type:

numpy.ndarray 2D dimension with (N,3) shape, where N is number of input coordinates

los_extrema(row, col, alt_min, alt_max)[source]

compute los extrema

Parameters:

  • row (float) – line sensor position

  • col (float) – column sensor position

  • alt_min (float) – los alt min

  • alt_max (float) – los alt max

Returns:

los extrema

Return type:

numpy.array (2x3)

interpolate_grid_in_plani(row, col)[source]

interpolate positions on multi h grid

Parameters:

  • row (float) – line sensor position

  • col (float) – column sensor position

Returns:

interpolated positions

Return type:

list

interpolate_grid_in_altitude(nbrow, nbcol, nbalt=None)[source]

interpolate equally spaced grid (in altitude)

Parameters:

  • nbrow (int) – grid nb row

  • nbcol (int) – grid nb col

  • nbalt (int) – grid nb alt, of None self.nbalt is used instead

Returns:

equally spaced grid

Return type:

numpy.array

direct_loc_grid_dtm(row0, col0, steprow, stepcol, nbrow, nbcol, dtm)[source]

direct localization grid on dtm

Parameters:

  • row0 (int) – grid origin (row)

  • col0 (int) – grid origin (col)

  • steprow (int) – grid step (row)

  • stepcol (int) – grid step (col)

  • nbrow (int) – grid nb row

  • nbcol (int) – grid nb col

  • dtm (shareloc.dtm) – dtm model

Returns:

direct localization grid

Return type:

numpy.array

return_grid_index(alt)[source]

return layer index enclosing a given altitude

Parameters:

alt (float) – altitude

Returns:

grid index (up,down)

Return type:

tuple

direct_loc_grid_h(row0, col0, steprow, stepcol, nbrow, nbcol, alt)[source]

direct localization grid at constant altitude TODO not tested.

Parameters:

  • row0 (int) – grid origin (row)

  • col0 (int) – grid origin (col)

  • steprow (int) – grid step (row)

  • stepcol (int) – grid step (col)

  • nbrow (int) – grid nb row

  • nbcol (int) – grid nb col

  • alt (float) – altitude of the grid

Returns:

direct localization grid

Return type:

numpy.array

estimate_inverse_loc_predictor(nbrow_pred=3, nbcol_pred=3)[source]

initialize inverse localization polynomial predictor it composed of 4 polynoms estimated on a grid at hmin and hmax

col_min = a0 + a1*lon + a2*lat + a3*lon**2 + a4*lat**2 + a5*lon*lat row_min = b0 + b1*lon + b2*lat + b3*lon**2 + b4*lat**2 + b5*lon*lat col_max = a0 + a1*lon + a2*lat + a3*lon**2 + a4*lat**2 + a5*lon*lat row_max = b0 + b1*lon + b2*lat + b3*lon**2 + b4*lat**2 + b5*lon*lat least squarred method is used to calculate coefficients, which are noramlized in [-1,1]

Parameters:

  • nbrow_pred (int) – predictor nb row (3 by default)

  • nbcol_pred (int) – predictor nb col (3 by default)

inverse_loc_predictor(lon, lat, alt=0.0)[source]

evaluate inverse localization predictor at a given geographic position

Parameters:

  • lon (float) – longitude

  • lat (float) – latitude

  • alt (float) – altitude (0.0 by default)

Returns:

sensor position and extrapolation state (row,col, is extrapolated)

Return type:

tuple (float, float, boolean)

inverse_partial_derivative(row, col, alt=0)[source]

calculate partial derivative at a given geographic position it gives the matrix to apply to get sensor shifts from geographic ones it returns M matrix : [dcol,drow]T = M x [dlon,dlat]T dlon/dlat in microrad M is calculated on each node of grid M is necessary for direct localization inversion in iterative inverse loc

Parameters:

  • lon (float) – longitude

  • lat (float) – latitude

  • alt (float) – altitude (0.0 by default)

Returns:

matrix

Return type:

numpy.array

inverse_loc(lon, lat, alt=0.0, nb_iterations=15)[source]

Inverse localization at a given geographic position First initialize position, * apply inverse predictor lon,lat,at -> col_0,row_0 * direct loc col_0,row_0 -> lon_0, lat_0 Then iterative process: * calculate geographic error dlon,dlat * calculate senor correction dlon,dlat -> dcol,drow * apply direct localization -> lon_i,lat_i

TODO explain algo TODO optimization (for loop,…)

Parameters:

  • lon (float) – longitude

  • lat (float) – latitude

  • alt (float) – altitude

  • nb_iterations (int) – max number of iterations (15 by default)

Returns:

sensor position (row,col,alt)

Return type:

tuple(1D np.array row position, 1D np.array col position, 1D np.array alt)

shareloc.geomodels.grid.coloc(multi_h_grid_src, multi_h_grid_dst, dtm, origin, step, size)[source]

colocalization grid on dtm localization on dtm from src grid, then inverse localization in right grid

Parameters:

  • multi_h_grid_src (shareloc.grid) – source grid

  • multi_h_grid_dst (shareloc.grid) – destination grid

  • origin (list(int)) – grid origin in src grid (row,col)

  • step (list(int)) – grid step (row,col)

  • size (list(int)) – grid nb row and nb col

Returns:

colocalization grid

Return type:

numpy.array