def main(manikin_name, vehicle_obj, settings):
"""
Perform all tasks related to intersection check between the specified vehicle and manikin
"""
for ang in settings["manikin_orientations"]:
ang %= 360.
manikin_folder = r"voxelated_models\manikins\{}\{}".format(
manikin_name, settings["voxel_size"])
manikin_filename = "voxels_{}_vox{}_ang{}".format(manikin_name, settings["voxel_size"],
ang)
logging.debug("Processing intersection between {} and {}".format(settings["run_id"],
manikin_name))
mani_data = data_io.load_array(manikin_folder, manikin_filename, multi_array=True)
manikin = Component(mani_data,
voxel_size=settings["voxel_size"])
################################################
# Process manikin
collisions = intersection_check(vehicle_obj, manikin, settings)
basename = "{}_{}".format(settings["run_id"], manikin_name)
result_folder = "results/{}/{}".format(basename, settings["voxel_size"])
file_name = "{}_vox{}_ang{}".format(basename, settings["voxel_size"], ang)
save_results(result_folder, file_name, collisions, save_plots=settings["show_2d"])
def main(manikin_name, vehicle_obj, settings):
"""
Perform all tasks related to intersection check between the specified vehicle and manikin
"""
for ang in settings["manikin_orientations"]:
ang %= 360.
manikin_folder = r"voxelated_models\manikins\{}\{}".format(
manikin_name, settings["voxel_size"])
manikin_filename = "voxels_{}_vox{}_ang{}".format(
manikin_name, settings["voxel_size"], ang)
logging.debug("Processing intersection between {} and {}".format(
settings["run_id"], manikin_name))
mani_data = data_io.load_array(manikin_folder,
manikin_filename,
multi_array=True)
manikin = Component(mani_data, voxel_size=settings["voxel_size"])
################################################
# Process manikin
collisions = intersection_check(vehicle_obj, manikin, settings)
basename = "{}_{}".format(settings["run_id"], manikin_name)
result_folder = "results/{}/{}".format(basename,
settings["voxel_size"])
file_name = "{}_vox{}_ang{}".format(basename, settings["voxel_size"],
ang)
save_results(result_folder,
file_name,
collisions,
save_plots=settings["show_2d"])
def load_create_graph(settings):
"""
Load previously calculated arrays of walkability scores and create graph structure based on
node connectivity
"""
result_folder = os.path.join("results", settings["run_id"], str(settings["voxel_size"]))
file_name = "{}_vox{}_best_scores".format(settings["run_id"], settings["voxel_size"])
overall_walkability_scores = data_io.load_array(result_folder, file_name)
file_name = "{}_vox{}_best_scores_by_orientation".format(settings["run_id"],
settings["voxel_size"])
scores_by_orie = data_io.load_array(result_folder,
file_name,
multi_array=True)
logging.debug("# nodes in unfiltered list: {}".format(overall_walkability_scores.size))
connected_points_graph = grid_to_graph_new(scores_by_orie, settings)
return connected_points_graph, overall_walkability_scores
def load_create_graph(settings):
"""
Load previously calculated arrays of walkability scores and create graph structure based on
node connectivity
"""
result_folder = os.path.join("results", settings["run_id"],
str(settings["voxel_size"]))
file_name = "{}_vox{}_best_scores".format(settings["run_id"],
settings["voxel_size"])
overall_walkability_scores = data_io.load_array(result_folder, file_name)
file_name = "{}_vox{}_best_scores_by_orientation".format(
settings["run_id"], settings["voxel_size"])
scores_by_orie = data_io.load_array(result_folder,
file_name,
multi_array=True)
logging.debug("# nodes in unfiltered list: {}".format(
overall_walkability_scores.size))
connected_points_graph = grid_to_graph_new(scores_by_orie, settings)
return connected_points_graph, overall_walkability_scores
def load_fit_of_pose_in_vehicle(basename, settings):
"""
Load in intersection check results for a given pose in the vehicle
:param basename: The manikin + vehicle name pair
:param settings:
:return: Dict of {angle: collision array} pairs for every orientation of mani in vehicle
"""
result_folder = "results/{}/{}".format(basename, settings["voxel_size"])
collision_arrays_dict = {}
for ang in settings["manikin_orientations"]:
data_file_name = "{}_vox{}_ang{}".format(basename, settings["voxel_size"], ang)
collision_arrays_dict[ang] = data_io.load_array(result_folder, data_file_name)
# Also combine results and store overall best fit (aka "does pose ever fit in this point?")
collision_arrays_dict["overall"] = find_best_scored_pose(
[collision_arrays_dict[k] for k in collision_arrays_dict])
return collision_arrays_dict
filemode="w",
format="%(levelname)s %(asctime)s %(message)s",
level=logging.DEBUG)
console = logging.StreamHandler()
console.setLevel(logging.INFO)
console.setFormatter(logging.Formatter("%(asctime)s %(message)s"))
logging.getLogger("").addHandler(console)
##########
# Load the pre-voxelated vehicle model for testing
b_fldr = r"voxelated_models/vehicles"
vox_size_str = str(SETTINGS["voxel_size"])
vox_folder = os.path.join(b_fldr, SETTINGS["run_id"], vox_size_str)
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"], vox_size_str)
voxel_dat = data_io.load_array(vox_folder, vox_veh_file, True)
vehicle = Vehicle(voxel_dat)
### Hardcoded dummy troop positions that may be meaningless. Uses 3D voxel indices. These
# values are for testing only, based on the Ricardo master assembly 11/11/2013.
crew_points = np.array([[40, 40, 25], [40, 80, 25], [20, 145, 40]])
# Perform exit checks and return combined results
can_exit, exit_grid, air_inside = main(vehicle, crew_points, debug=True)
# Print exit status: do all points pass exit check?
exit_ok = all(v.values() for v in can_exit.values())
if SETTINGS["debug"]:
## Render a 3D image, for debugging purposes only
from mayavi import mlab
if __name__ == "__main__":
# User specified settings from separate file
from rpl.tools.api import test_bench_api as tba
SETTINGS = tba.load_settings("settings.js")
## Set up logging
logging.info("\n"+50*"_"+'\ntest bench started\n'+50*"_")
# Value references numpy; can't be read from json
SETTINGS["unwalkable_value"] = np.inf
# Load in vehicle occupied voxels file in order to locate the door points
vox_veh_folder = r"voxelated_models/vehicles/{}/{}".format(SETTINGS["run_id"],
SETTINGS["voxel_size"])
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"], SETTINGS["voxel_size"])
veh_vox = data_io.load_array(vox_veh_folder, vox_veh_file, multi_array=True)
#### Standalone egress testing mode requires a few dummy points that will depend on the
# assembly chosen. These apply to the Ricardo master assembly as of Nov 11, 2013.
vox_veh_cabin = (29, 63, 20) # Dummy position; must label floor regions to find door-adj
## Point chosen for the EFV tester model
#vox_veh_cabin = (140,30,10)
#### Some sample output.
# Gridpoints for 5cm voxel spacings. One point from cabin (exit exists), and one point for
# driver ( can not reach door)
start_points = [(29, 63, 4), (46, 45, 4), (29, 30, 4), (46, 67, 4), (29, 74, 4), (46, 34, 4),
(29, 52, 4), (46, 78, 4), (46, 23, 4), (29, 96, 4), (29, 85, 4), (46, 56, 4),
(46, 12, 4), (29, 19, 4), (29, 41, 4), (46, 89, 4)]
#######
# Voxelize the vehicle and return all information
# Save the voxelated model of the vehicle (sans door and other excluded parts)
vox_veh_folder = r"voxelated_models/vehicles/{}/{}".format(
SETTINGS["run_id"], SETTINGS["voxel_size"])
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"],
SETTINGS["voxel_size"])
if SETTINGS["debug"] and os.path.exists(
os.path.join(vox_veh_folder, vox_veh_file + '.npz')):
# Load from pre-existing file if possible (if debugging option is on)
assembly.voxel_data = data_io.load_array(vox_veh_folder,
vox_veh_file,
multi_array=True)
logging.info(
"Vehicle voxel data loaded from preexisting file {}".format(
vox_veh_file))
valid_litter_found = "litter" in assembly.voxel_data # Does cached geom incl litter?
else:
logging.info('Commencing voxelization of vehicle model')
assembly.voxelate_geometry()
# Voxelate Litter_Open model on a grid of same size
x_pts, y_pts, z_pts = [
assembly.voxel_data[k] for k in ("x_grid", "y_grid", "z_grid")
]
valid_litter_found = assembly.voxelate_litter(x_pts, y_pts, z_pts)
#######
# Call test bench api to obtain desired information
assembly = Assembly_Info(SETTINGS)
#######
# Voxelize the vehicle and return all information
# Save the voxelated model of the vehicle (sans door and other excluded parts)
vox_veh_folder = r"voxelated_models/vehicles/{}/{}".format(SETTINGS["run_id"],
SETTINGS["voxel_size"])
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"], SETTINGS["voxel_size"])
if SETTINGS["debug"] and os.path.exists(os.path.join(vox_veh_folder, vox_veh_file + '.npz')):
# Load from pre-existing file if possible (if debugging option is on)
assembly.voxel_data = data_io.load_array(vox_veh_folder, vox_veh_file, multi_array=True)
logging.info("Vehicle voxel data loaded from preexisting file {}".format(vox_veh_file))
valid_litter_found = "litter" in assembly.voxel_data # Does cached geom incl litter?
else:
logging.info('Commencing voxelization of vehicle model')
assembly.voxelate_geometry()
# Voxelate Litter_Open model on a grid of same size
x_pts, y_pts, z_pts = [assembly.voxel_data[k] for k in ("x_grid", "y_grid", "z_grid")]
valid_litter_found = assembly.voxelate_litter(x_pts, y_pts, z_pts)
data_io.save_multi_array(vox_veh_folder, vox_veh_file, assembly.voxel_data)
logging.info("Vehicle successfully voxelated.")
# All later operations act on pre-voxelated models. Create voxel objects with redundant info.
filemode="w",
format="%(levelname)s %(asctime)s %(message)s",
level=logging.DEBUG)
console = logging.StreamHandler()
console.setLevel(logging.INFO)
console.setFormatter(logging.Formatter("%(asctime)s %(message)s"))
logging.getLogger("").addHandler(console)
##########
# Load the pre-voxelated vehicle model for testing
b_fldr = r"voxelated_models/vehicles"
vox_size_str = str(SETTINGS["voxel_size"])
vox_folder = os.path.join(b_fldr, SETTINGS["run_id"], vox_size_str)
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"], vox_size_str)
voxel_dat = data_io.load_array(vox_folder, vox_veh_file, True)
vehicle = Vehicle(voxel_dat)
### Hardcoded dummy troop positions that may be meaningless. Uses 3D voxel indices. These
# values are for testing only, based on the Ricardo master assembly 11/11/2013.
crew_points = np.array([[40, 40, 25], [40, 80, 25], [20, 145, 40]])
# Perform exit checks and return combined results
can_exit, exit_grid, air_inside = main(vehicle, crew_points, debug=True)
# Print exit status: do all points pass exit check?
exit_ok = all(v.values() for v in can_exit.values())
if SETTINGS["debug"]:
## Render a 3D image, for debugging purposes only
from mayavi import mlab
from rpl.tools.api import test_bench_api as tba
SETTINGS = tba.load_settings("settings.js")
## Set up logging
logging.info("\n" + 50 * "_" + '\ntest bench started\n' + 50 * "_")
# Value references numpy; can't be read from json
SETTINGS["unwalkable_value"] = np.inf
# Load in vehicle occupied voxels file in order to locate the door points
vox_veh_folder = r"voxelated_models/vehicles/{}/{}".format(
SETTINGS["run_id"], SETTINGS["voxel_size"])
vox_veh_file = "voxels_{}_vox{}".format(SETTINGS["run_id"],
SETTINGS["voxel_size"])
veh_vox = data_io.load_array(vox_veh_folder,
vox_veh_file,
multi_array=True)
#### Standalone egress testing mode requires a few dummy points that will depend on the
# assembly chosen. These apply to the Ricardo master assembly as of Nov 11, 2013.
vox_veh_cabin = (
29, 63, 20
) # Dummy position; must label floor regions to find door-adj
## Point chosen for the EFV tester model
#vox_veh_cabin = (140,30,10)
#### Some sample output.
# Gridpoints for 5cm voxel spacings. One point from cabin (exit exists), and one point for
# driver ( can not reach door)
start_points = [(29, 63, 4), (46, 45, 4), (29, 30, 4), (46, 67, 4),
