def atlaas_grid(filepath, var_threshold=0.1, stamp=None, frame_id='/map'):
""" Usage:
atlaas.merge("atlaas.*x*.tif", "out.tif")
if os.path.isfile("out.tif"):
atlaas_grid("out.tif")
"""
g = gdal2(filepath)
bnp = g.bands[g.names['N_POINTS']]
bma = g.bands[g.names['Z_MAX']]
bme = g.bands[g.names['Z_MEAN']]
bva = g.bands[g.names['VARIANCE']]
data = numpy.where(bva > var_threshold, bma, bme)
delta = data.max() - data.min()
ddis = (100.0 * (data - data.min()) / delta).astype('uint8')
ddis[bnp < 1] = 0 # no points, unknown
og = OccupancyGrid()
og.data = ddis.flatten()
if stamp: og.header.stamp = stamp
og.header.frame_id = frame_id
og.info.resolution = g.scale_x
og.info.height, og.info.width = data.shape
og.info.origin.position.x, og.info.origin.position.y = g.u2c(
g.utm_x, g.utm_y)
og.info.origin.orientation.x = 1 # flip to transform UTM-ROS (scale_y < 0)
return og
def atlaas_grid(filepath, var_threshold=0.1, stamp=None, frame_id="/map"):
""" Usage:
atlaas.merge("atlaas.*x*.tif", "out.tif")
if os.path.isfile("out.tif"):
atlaas_grid("out.tif")
"""
g = gdal2(filepath)
bnp = g.bands[g.names["N_POINTS"]]
bma = g.bands[g.names["Z_MAX"]]
bme = g.bands[g.names["Z_MEAN"]]
bva = g.bands[g.names["VARIANCE"]]
data = numpy.where(bva > var_threshold, bma, bme)
delta = data.max() - data.min()
ddis = (100.0 * (data - data.min()) / delta).astype("uint8")
ddis[bnp < 1] = 0 # no points, unknown
og = OccupancyGrid()
og.data = ddis.flatten()
if stamp:
og.header.stamp = stamp
og.header.frame_id = frame_id
og.info.resolution = g.scale_x
og.info.height, og.info.width = data.shape
og.info.origin.position.x, og.info.origin.position.y = g.u2c(g.utm_x, g.utm_y)
og.info.origin.orientation.x = 1 # flip to transform UTM-ROS (scale_y < 0)
return og
def convert(fin, fout, cmin, cmax, cmap="viridis"):
geo = gdal2(fin)
npt = geo.bands[geo.names["N_POINTS"]]
img = geo.bands[geo.names["Z_MEAN"]]
img = (img - cmin) * (1.0 / (cmax - cmin))
img[img > 1] = 1
img[img < 0] = 0
cmap = getattr(cm, cmap)
img = (cmap(img) * 255).astype("uint8")
img[npt < 1] = 0
save(fout, img)
def convert(fin, fout, cmin, cmax, cmap=colormap, npt=None):
geo = gdal2(fin)
img = geo.bands
if len(img.shape) > 2: # multi-layer
npt = geo.bands[geo.names["N_POINTS"]]
img = geo.bands[geo.names["Z_MEAN"]]
img = (img - cmin) * (1./(cmax - cmin))
img[img > 1] = 1
img[img < 0] = 0
img[npt < 1] = 0
# in case of JPEG or WebP, set quality to 90%, else this option is ignored
save(fout, (cmap(img)*255).astype('uint8'))
def convert(fin, fout, cmin, cmax, cmap=colormap, npt=None):
geo = gdal2(fin)
img = geo.bands
if len(img.shape) > 2: # multi-layer
npt = geo.bands[geo.names["N_POINTS"]]
img = geo.bands[geo.names["Z_MEAN"]]
img = (img - cmin) * (1. / (cmax - cmin))
img[img > 1] = 1
img[img < 0] = 0
img[npt < 1] = 0
# in case of JPEG or WebP, set quality to 90%, else this option is ignored
save(fout, (cmap(img) * 255).astype('uint8'))
def atlaas8u_grid(filepath, stamp=None, frame_id='/map'):
g = gdal2(filepath)
b = g.bands[0] if len(g.bands.shape) > 2 else g.bands
og = OccupancyGrid()
og.data = (b.astype('float') / 2.55).astype('uint8').flatten()
if stamp: og.header.stamp = stamp
og.header.frame_id = frame_id
og.info.resolution = g.scale_x
og.info.height, og.info.width = b.shape
og.info.origin.position.x, og.info.origin.position.y = g.u2c(
g.utm_x, g.utm_y)
og.info.origin.orientation.x = 1 # flip to transform UTM-ROS (scale_y < 0)
return og
def atlaas8u_grid(filepath, stamp=None, frame_id="/map"):
g = gdal2(filepath)
b = g.bands[0] if len(g.bands.shape) > 2 else g.bands
og = OccupancyGrid()
og.data = (b.astype("float") / 2.55).astype("uint8").flatten()
if stamp:
og.header.stamp = stamp
og.header.frame_id = frame_id
og.info.resolution = g.scale_x
og.info.height, og.info.width = b.shape
og.info.origin.position.x, og.info.origin.position.y = g.u2c(g.utm_x, g.utm_y)
og.info.origin.orientation.x = 1 # flip to transform UTM-ROS (scale_y < 0)
return og
import os
import json
from PIL import Image, ImageDraw
from atlaas.helpers.gdal2 import gdal2
with open('poses.json') as f:
poses = json.load(f)
g = gdal2('merged.atlaas.00000.tif')
xy_atlaas = [tuple(map(int, g.c2p(x,y))) for x,y in poses]
inc = 0
while os.path.isfile('movie4atlaas.%05i.png'%inc):
im = Image.open('movie4atlaas.%05i.png'%inc)
dr = ImageDraw.Draw(im)
dr.line(xy_atlaas[:inc+1], fill=(255,0,0,255)) # , width=2)
im.save('movie4atlaas.path.%05i.png'%inc)
inc += 1
import os
import glob
import atlaas
from atlaas.helpers.gdal2 import gdal2
from atlaas.helpers.image import zero
if not glob.glob("atlaas.*x*.tif"):
test = atlaas.Atlaas()
test.init(120.0, 120.0, 0.1, 377016.5, 4824342.9, 141.0, 31, True)
test.process()
test.save_currents()
del test
# 1st erase data in tiles so as to have same size in merge() without noise
for fatlaas in glob.glob("atlaas.*x*.tif"):
g = gdal2(fatlaas)
g.bands *= 0
g.save()
fregion = "region.%s.png"%fatlaas.split('.')[1]
atlaas.tile_to_region(fatlaas, fregion)
for fregion in glob.glob("region.*x*.png"):
zero(fregion) # empty data
pcd_id = 0
patern = 'cloud.%05i.pcd'
test = atlaas.Atlaas()
test.init(120.0, 120.0, 0.1, 377016.5, 4824342.9, 141.0, 31, True)
while os.path.isfile(patern%pcd_id):
# coding: utf-8 get_ipython().magic(u'matplotlib inline') filename = 'out.tif' from atlaas.helpers.gdal2 import gdal2 from atlaas.helpers.matplot import show, hist g = gdal2(filename) # <codecell> band = g.bands[g.names['N_POINTS']] hist(band) # <codecell> show(band, 0, 100) # <codecell> band = g.bands[g.names['VARIANCE']] hist(band) # <codecell> show(band, 0, 0.2) # <codecell> band = g.bands[g.names['Z_MEAN']]
import os
import json
from PIL import Image, ImageDraw
from atlaas.helpers.gdal2 import gdal2
with open('poses.json') as f:
poses = json.load(f)
g = gdal2('merged.region.00000.png')
xy_region = [tuple(map(int, g.c2p(x,y))) for x,y in poses]
inc = 0
while os.path.isfile('movie4region.%05i.png'%inc):
im = Image.open('movie4region.%05i.png'%inc)
dr = ImageDraw.Draw(im)
dr.line(xy_region[:inc+1], fill=(255,0,0,255))
im.save('movie4region.path.%05i.png'%inc)
inc += 1
