gdalwrap::gdal dsm_to_region(const gdalwrap::gdal& dsm) {
gdalwrap::gdal region;
region.copy_meta(dsm, clara::region::N_RASTER);
const auto& dsm_band = dsm.bands[0];
size_t pose, width = dsm.get_width();
for (size_t px_x = 0; px_x < width - 1; px_x++)
for (size_t px_y = 0; px_y < dsm.get_height() - 1; px_y++) {
// compute :: Z_MEAN{x1 - x2} > 30cm : P_OBSTACLE = 1
pose = px_x + px_y * width;
flag_obstacle(dsm_band, region.bands, pose, pose + 1);
flag_obstacle(dsm_band, region.bands, pose, pose + width);
flag_obstacle(dsm_band, region.bands, pose, pose + width + 1);
}
region.names = {"NO_3D_CLASS", "FLAT", "OBSTACLE", "ROUGH", "SLOPE"};
return region;
}
static bpy::list py_get_band_as_uchar(gdalwrap::gdal& self, const std::string& name) {
return std_vector_to_py_list(gdalwrap::raster2bytes(self.get_band(name)));
}
static bpy::list py_get_band(gdalwrap::gdal& self, const std::string& name) {
return std_vector_to_py_list(self.get_band(name));
}
void save(const std::string& filepath) const {
io.save(filepath);
}
void save(const std::string& filepath) {
dtm.save(filepath);
}
const gdalwrap::raster& get_npointsmap() const {
return dtm.get_band("N_POINTS");
}
/* getters */
const gdalwrap::raster& get_heightmap() const {
return dtm.get_band("Z_MAX");
}
/** Get the index of the point in the raster
*
* @param p the point
*
* @returns the index of the point
*
*/
size_t index( const point_xy_t& p ) const {
return dtm.index_custom(p[0], p[1]);
}
/** load region and robot model
*
* @param f_region path to a region.tif file
* (multi-layers terrains classification probabilities, float32)
*
* @param f_robot_model path to a robot model
* to generate the visibility map (at least its sensor height)
*
*/
void load(const std::string& f_dtm, const std::string& f_robot_model) {
dtm.load(f_dtm);
rmdl.load(f_robot_model);
_load();
}