def copy_top_match(querydir, query, ranked_matches, match):
topentry = ranked_matches[0]
matchedimg = topentry[0]
score = topentry[1]
dup = "dup" + str(len(ranked_matches) == 1 or score == ranked_matches[1][1])
qlat = float(query.split(',')[1])
qlon = float(query.split(',')[2])
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
queryimgpath = os.path.join(querydir, query + '.pgm')
queryoutpath = os.path.join(resultsdir, query + ';query;gt' + str(match) + ';' + dup + ';' + matchedimg + ';' + str(score) + ';' + str(distance) + '.pgm')
shutil.copyfile(queryimgpath, queryoutpath)
for matchedimg, score in ranked_matches:
if score != topentry[1]:
break
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
matchimgpath = os.path.join(dbdump, '%s.jpg' % matchedimg)
matchoutpath = os.path.join(resultsdir, query + ';match;gt' + str(match) + ';' + dup + ';' + matchedimg + ';' + str(score) + ';' + str(distance) + '.jpg')
shutil.copy(matchimgpath, matchoutpath)
def copy_top_match(querydir, query, ranked_matches, match):
topentry = ranked_matches[0]
matchedimg = topentry[0]
score = topentry[1]
dup = "dup" + str(
len(ranked_matches) == 1 or score == ranked_matches[1][1])
qlat = float(query.split(',')[1])
qlon = float(query.split(',')[2])
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
queryimgpath = os.path.join(querydir, query + '.pgm')
queryoutpath = os.path.join(
resultsdir, query + ';query;gt' + str(match) + ';' + dup + ';' +
matchedimg + ';' + str(score) + ';' + str(distance) + '.pgm')
shutil.copyfile(queryimgpath, queryoutpath)
for matchedimg, score in ranked_matches:
if score != topentry[1]:
break
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
matchimgpath = os.path.join(dbdump, '%s.jpg' % matchedimg)
matchoutpath = os.path.join(
resultsdir, query + ';match;gt' + str(match) + ';' + dup + ';' +
matchedimg + ';' + str(score) + ';' + str(distance) + '.jpg')
shutil.copy(matchimgpath, matchoutpath)
def lltom(lat1, lon1, lat2, lon2):
"""Returns lat2,lon2 relative to lat1,lon1 in meters.
Does not handle wraparound at 360."""
if lat1 == lat2 and lon1 == lon2:
return [0, 0]
dx = distance(lat1, lon1, lat2, lon1) * (-1 if lat2 < lat1 else 1)
dy = distance(lat1, lon1, lat1, lon2) * (-1 if lon2 < lon1 else 1)
return [dx, dy]
def generateCellCoords(lat, lon, len, distance):
#takes lat long as upper left corner and length as distance to go right and down 210 degrees
#distance is distance between cells
coords = []
candlat1 = lat
candlon1 = lon
while(info.distance(lat, lon, candlat1, candlon1) < len):
candlat2, candlon2 = candlat1, candlon1
while(info.distance(candlat1, candlon1, candlat2, candlon2) < len):
coords.append((candlat2, candlon2))
candlat2, candlon2 = em.moveLocation4(candlat2, candlon2, distance, 90)
candlat1, candlon1 = em.moveLocation4(candlat1, candlon1, distance, 210)
return coords
def hasView(self, C, lat, lon, qlat, qlon, qyaw, thresh):
dist = distance(lat, lon, qlat, qlon)
rad = min(20, max(3, int(dist/10)))
def similar(view):
return distance(view.lat, view.lon, qlat, qlon) < thresh
def yawof(view):
return view.yaw
views = list(self.getViews(C, lat, lon, rad))
# print "dist is %d, rad is %d, nviews is %d" % (dist, rad, len(views))
closeyaws = map(lambda (k,v): v, sorted([(distance(v.lat, v.lon, qlat, qlon), yawof(v)) for v in views]))[:10]
norm = geom.compute_norm(closeyaws)
return any([similar(v) for v in views]), norm
def taggedcopy(self, points, image, correction=False):
MIN_SIZE = 1
draw = ImageDraw.Draw(image)
def dist(p):
return info.distance(p[0].lat, p[0].lon, self.lat, self.lon)
points.sort(key=dist, reverse=True) # draw distant first
if not correction:
points = [(t,(d,p),1) for (t,(d,p)) in points]
for tag, (dist, point), correction in points:
color = self.colordist(dist, 30.0)
size = int(300.0*correction/info.distance(tag.lat, tag.lon, self.lat, self.lon))
fontPath = "/usr/share/fonts/truetype/ttf-dejavu/DejaVuSans-Bold.ttf"
font = ImageFont.truetype(fontPath, max(size, MIN_SIZE))
off_x = -size*2
off_y = -size*(len(tag)+1)
# start black container
top_left = (point[0] + off_x - 3, point[1] + off_y - 1)
w = 10
for line in tag:
w = max(w, draw.textsize(line, font)[0])
bottom_right = (point[0] + off_x + w + 3, point[1] + off_y + max(size, MIN_SIZE)*len(tag) + 3)
img = image.copy()
draw2 = ImageDraw.Draw(img)
draw2.rectangle([top_left, bottom_right], fill='#000')
image = Image.blend(image, img, 0.75)
draw = ImageDraw.Draw(image)
# end black container
draw.ellipse((point[0]-size/2,point[1]-size/2,point[0]+size/2,point[1]+size/2), fill=color)
for line in tag:
draw.text((point[0] + off_x, point[1] + off_y), line, fill=color, font=font)
off_y += max(size, MIN_SIZE)
# if dist:
# INFO('mapping tag at %f meters error' % dist)
return image
def admit(match):
line = match[0][:-8]
lat, lon = getlatlonfromdbimagename(C, line)
dist = info.distance(lat, lon, Q.query_lat, Q.query_lon)
if dist < C.amb_cutoff + C.amb_padding:
return True
return False
def query2(querydir, querysurf, dbdir, mainOutputDir, nClosestCells, copytopmatch, params, copy_top_n_percell=0):
lat, lon = info.getQuerySURFCoord(querysurf)
closest_cells = util.getclosestcells(lat, lon, dbdir)
assert dict(closest_cells)['37.8732916331,-122.268346029'] < 236 or dict(closest_cells)['37.8714489427,-122.272389514'] < 236 or dict(closest_cells)['37.8696062215,-122.273737308'] < 236
built = [
'37.8732916331,-122.268346029',
'37.8714489427,-122.272389514',
'37.8696062215,-122.273737308',
]
closest_cells = filter(lambda c: c[0] in built, closest_cells)
outputFilePaths = []
cells_in_range = [(cell, dist) for cell, dist in closest_cells[0:nClosestCells] if dist < cellradius + ambiguity+matchdistance]
latquery, lonquery = info.getQuerySURFCoord(querysurf)
if verbosity > 0:
print "checking query: {0} \t against {1} \t cells".format(querysurf, len(cells_in_range))
for cell, dist in cells_in_range:
latcell, loncell = cell.split(',')
latcell = float(latcell)
loncell = float(loncell)
actualdist = info.distance(latquery, lonquery, latcell, loncell)
if verbosity > 1:
print "querying cell: {0}, distance: {1} with:{2}".format(cell, actualdist, querysurf)
outputFilePath = os.path.join(mainOutputDir, querysurf + ',' + cell + ',' + str(actualdist) + ".res")
outputFilePaths.append(outputFilePath)
# start query
query.run_parallel(dbdir, [c for c,d in cells_in_range], querydir, querysurf, outputFilePaths, params)
# end query
for cell, dist in cells_in_range:
latcell, loncell = cell.split(',')
latcell = float(latcell)
loncell = float(loncell)
actualdist = info.distance(latquery, lonquery, latcell, loncell)
outputFilePath = os.path.join(mainOutputDir, querysurf + ',' + cell + ',' + str(actualdist) + ".res")
if copy_top_n_percell > 0:
outputDir = os.path.join(mainOutputDir, querysurf + ',' + cell + ',' + str(actualdist))
copy_topn_results(os.path.join(dbdir, cell), outputDir, outputFilePath, 4)
# combined = combine_until_dup(outputFilePaths, 1000)
combined = combine_topn_votes(outputFilePaths, float('inf'))
# combined = filter_in_range(combined, querysurf)
# write_scores(querysurf, combined, "/media/data/combined")
[g, y, r, b, o] = check_topn_img(querysurf, combined, topnresults)
if copytopmatch:
match = g or y or r or b or o
copy_top_match(querydir, querysurf.split('surf.npy')[0], combined, match)
return [g, y, r, b, o]
def hasView(self, C, lat, lon, qlat, qlon, qyaw, thresh):
dist = distance(lat, lon, qlat, qlon)
rad = min(20, max(3, int(dist / 10)))
def similar(view):
return distance(view.lat, view.lon, qlat, qlon) < thresh
def yawof(view):
return view.yaw
views = list(self.getViews(C, lat, lon, rad))
# print "dist is %d, rad is %d, nviews is %d" % (dist, rad, len(views))
closeyaws = map(
lambda (k, v): v,
sorted([(distance(v.lat, v.lon, qlat, qlon), yawof(v))
for v in views]))[:10]
norm = geom.compute_norm(closeyaws)
return any([similar(v) for v in views]), norm
def visualize_point_cloud(self, cloudfile, output):
cloud = np.load(cloudfile).item()
img = self.image.copy()
pixels = img.load()
for x,y in cloud:
d = cloud[x,y]
lat, lon = d['lat'], d['lon']
pixels[x,y] = self.colordist(info.distance(lat, lon, self.lat, self.lon))
img = Image.blend(img, self.image, 0.5)
img.save(output, 'png')
def skew_location(C, Q):
length = 2*C.ambiguity
points = []
corner = info.moveLocation(Q.sensor_lat, Q.sensor_lon, (2**.5)*C.ambiguity, -45)
for i in range(length+1):
row = info.moveLocation(corner[0], corner[1], i, 180)
for j in range(length+1):
point = info.moveLocation(row[0], row[1], j, 90)
if info.distance(Q.sensor_lat, Q.sensor_lon, point[0], point[1]) <= C.ambiguity:
points.append(point)
return points
def filter_in_range(ranked_matches, querysurf):
qlat, qlon = info.getQuerySURFCoord(querysurf)
weighted_matches = []
for matchedimg, score in ranked_matches:
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
if distance < ambiguity + matchdistance:
weighted_matches.append((matchedimg, 2 * score))
weighted_matches.sort(key=lambda x: x[1], reverse=True)
return weighted_matches
def filter_in_range(ranked_matches, querysurf):
qlat, qlon = info.getQuerySURFCoord(querysurf)
weighted_matches = []
for matchedimg, score in ranked_matches:
clat = float(matchedimg.split(",")[0])
clon = float(matchedimg.split(",")[1][0:-5])
distance = info.distance(qlat, qlon, clat, clon)
if distance < ambiguity+matchdistance:
weighted_matches.append((matchedimg, 2 * score))
weighted_matches.sort(key=lambda x: x[1], reverse=True)
return weighted_matches
def getNumJPGInRange(lat, lon, inputDir, radius, files):
"""counts all images in inputDir within radius of given coordinate makes assumption about format of filename"""
if os.path.exists(inputDir):
count = 0
for file in files:
lat2, lon2 = info.getImgCoord(str(file))
dist = info.distance(lat, lon, lat2, lon2)
if(dist < radius):
count+=1
return count
else:
raise OSError("{p} does not exist.".format(p=inputDir))
def skew_location(querysift, radius):
center = info.getQuerySIFTCoord(querysift)
length = 2*radius
points = []
corner = info.moveLocation(center[0], center[1], (2**.5)*radius, -45)
for i in range(length+1):
row = info.moveLocation(corner[0], corner[1], i, 180)
for j in range(length+1):
point = info.moveLocation(row[0],row[1], j, 90)
if info.distance(center[0],center[1], point[0], point[1]) <= radius:
points.append(point)
return points
def select_frustum(self, lat, lon, yaw, fov=270, radius=100):
contained = []
for tag in self.tags:
dist = info.distance(lat, lon, tag.lat, tag.lon)
if dist > radius:
continue
v1 = (math.sin(math.radians(yaw)), math.cos(math.radians(yaw)))
v2 = (tag.lon-lon, tag.lat-lat)/linalg.norm((tag.lon-lon, tag.lat-lat))
degrees = math.acos(np.dot(v1,v2))*180/math.pi
assert degrees > 0
if degrees > fov/2:
continue
contained.append(tag)
return contained
def copySIFTInRange(lat, lon, inputDir, outputDir, radius, files):
"""puts all sift files in inputDir within radius of given coordinate into outputDir.
makes assumption about format of filename"""
if not os.path.exists(outputDir):
try:
os.makedirs(outputDir)
except Exception:
print "Error making directory...quitting..."
return
for file in files:
lat2, lon2 = info.getSIFTCoord(str(file))
dist = info.distance(lat, lon, lat2, lon2)
if(dist < radius):
os.symlink(os.path.join(os.path.abspath(inputDir + '/'), file), os.path.join(outputDir + '/', file))
def select_frustum(self, lat, lon, yaw, fov=270, radius=100):
contained = []
for tag in self.tags:
dist = info.distance(lat, lon, tag.lat, tag.lon)
if dist > radius:
continue
v1 = (math.sin(math.radians(yaw)), math.cos(math.radians(yaw)))
v2 = (tag.lon - lon, tag.lat - lat) / linalg.norm(
(tag.lon - lon, tag.lat - lat))
degrees = math.acos(np.dot(v1, v2)) * 180 / math.pi
assert degrees > 0
if degrees > fov / 2:
continue
contained.append(tag)
return contained
def getclosestcell(lat, lon, celldir):
if not os.path.exists(celldir):
return
dirs = getdirs(celldir)
closest_dist = float('inf')
closest_cell = ''
for dir in dirs:
lat2, lon2 = dir.split(',')
lat2 = float(lat2)
lon2 = float(lon2)
dist = info.distance(lat, lon, lat2, lon2)
if dist < closest_dist:
closest_dist = dist
closest_cell = dir
return closest_cell, closest_dist
def getclosestcells(lat, lon, celldir):
if celldir in dircache:
dirs = dircache[celldir]
else:
if not os.path.exists(celldir):
print "ERR: celldir does not exist: {0}".format(celldir)
return []
dirs = dircache[celldir] = getdirs(celldir)
closest_cells = []
for dir in dirs:
lat2, lon2 = dir.split(',')
lat2 = float(lat2)
lon2 = float(lon2)
dist = info.distance(lat, lon, lat2, lon2)
closest_cells.append((dir, dist))
closest_cells.sort(key=lambda x: x[1])
return closest_cells
def norm_compatible(pt, viewpt, verbose=False):
# no bearing: corner, middle of street, etc
if pt.bearing is None:
return True
# really close up!
if distance(pt.lat, pt.lon, viewpt.lat, viewpt.lon) < 20:
return True
yaw = getbearing(pt.lat, pt.lon, viewpt.lat, viewpt.lon)
diff = anglediff(pt.bearing * pi / 180, yaw * pi / 180) * 180 / math.pi
if verbose:
print "Point", pt
print "pt bearing", pt.bearing
print "view bearing", yaw
print "diff", diff
print
return diff < 85
def filter_locally_significant(self, fgen, d0, m0, q, rev=False):
T = 70000 # pts ABOVE threshold allowed
R = 25.0 # max visibility distance
NN = 10 # neighbors to consider
flann = pyflann.FLANN()
INFO("building flann index")
q.flann_setup_index(flann, d0, m0, None)
INFO("building 3d map")
map3d = self.load_3dmap_for_cell(q.cellpath, d0, m0, q.infodir)
INFO("begin filtering features")
params = q.params.copy()
params['num_neighbors'] = NN
for offset, fchunk in fgen:
fchunk_filtered = []
arr = np.ndarray(len(fchunk), self.dtype)
arr[:] = fchunk
results, dists = flann.nn_index(arr['vec'], **params)
# for each feature
for _, (index_arr, dist_arr) in enumerate(zip(results, dists)):
refpt = map3d[offset + _]
# for each match of feature
has_r = False
ok = False
for i, d in zip(index_arr, dist_arr):
pt = map3d[i]
r = distance(pt['lat'], pt['lon'], refpt['lat'],
refpt['lon'])
if r > R:
has_r = True
if (not rev and d > T) or (rev and d < T):
ok = True
break
if ok or not has_r:
fchunk_filtered.append(fchunk[_])
print 'filter chunk %d/%d' % (len(fchunk_filtered), len(fchunk))
yield fchunk_filtered
def filter_locally_significant(self, fgen, d0, m0, q, rev=False):
T = 70000 # pts ABOVE threshold allowed
R = 25.0 # max visibility distance
NN = 10 # neighbors to consider
flann = pyflann.FLANN()
INFO("building flann index")
q.flann_setup_index(flann, d0, m0, None)
INFO("building 3d map")
map3d = self.load_3dmap_for_cell(q.cellpath, d0, m0, q.infodir)
INFO("begin filtering features")
params = q.params.copy()
params['num_neighbors'] = NN
for offset, fchunk in fgen:
fchunk_filtered = []
arr = np.ndarray(len(fchunk), self.dtype)
arr[:] = fchunk
results, dists = flann.nn_index(arr['vec'], **params)
# for each feature
for _, (index_arr, dist_arr) in enumerate(zip(results, dists)):
refpt = map3d[offset + _]
# for each match of feature
has_r = False
ok = False
for i, d in zip(index_arr, dist_arr):
pt = map3d[i]
r = distance(pt['lat'], pt['lon'], refpt['lat'], refpt['lon'])
if r > R:
has_r = True
if (not rev and d > T) or (rev and d < T):
ok = True
break
if ok or not has_r:
fchunk_filtered.append(fchunk[_])
print 'filter chunk %d/%d' % (len(fchunk_filtered), len(fchunk))
yield fchunk_filtered
def distance_f(oloc): return info.distance(source.lat, source.lon, oloc['lat'], oloc['lon'])
def similar(view):
return distance(view.lat, view.lon, qlat, qlon) < thresh
def xydistance(self, d2): x1, y1, z1 = self.lat, self.lon, self.alt x2, y2, z2 = d2['lat'], d2['lon'], d2['alt'] xydist = info.distance(x1, y1, x2, y2) return xydist
def distance_f(oloc):
return info.distance(source.lat, source.lon, oloc['lat'],
oloc['lon'])
def dist(p): return info.distance(p[0].lat, p[0].lon, self.lat, self.lon)
def xydistance(self, d2):
x1, y1, z1 = self.lat, self.lon, self.alt
x2, y2, z2 = d2['lat'], d2['lon'], d2['alt']
xydist = info.distance(x1, y1, x2, y2)
return xydist
def query2(querydir,
querysurf,
dbdir,
mainOutputDir,
nClosestCells,
copytopmatch,
params,
copy_top_n_percell=0):
lat, lon = info.getQuerySURFCoord(querysurf)
closest_cells = util.getclosestcells(lat, lon, dbdir)
assert dict(closest_cells)['37.8732916331,-122.268346029'] < 236 or dict(
closest_cells)['37.8714489427,-122.272389514'] < 236 or dict(
closest_cells)['37.8696062215,-122.273737308'] < 236
built = [
'37.8732916331,-122.268346029',
'37.8714489427,-122.272389514',
'37.8696062215,-122.273737308',
]
closest_cells = filter(lambda c: c[0] in built, closest_cells)
outputFilePaths = []
cells_in_range = [(cell, dist)
for cell, dist in closest_cells[0:nClosestCells]
if dist < cellradius + ambiguity + matchdistance]
latquery, lonquery = info.getQuerySURFCoord(querysurf)
if verbosity > 0:
print "checking query: {0} \t against {1} \t cells".format(
querysurf, len(cells_in_range))
for cell, dist in cells_in_range:
latcell, loncell = cell.split(',')
latcell = float(latcell)
loncell = float(loncell)
actualdist = info.distance(latquery, lonquery, latcell, loncell)
if verbosity > 1:
print "querying cell: {0}, distance: {1} with:{2}".format(
cell, actualdist, querysurf)
outputFilePath = os.path.join(
mainOutputDir,
querysurf + ',' + cell + ',' + str(actualdist) + ".res")
outputFilePaths.append(outputFilePath)
# start query
query.run_parallel(dbdir, [c for c, d in cells_in_range], querydir,
querysurf, outputFilePaths, params)
# end query
for cell, dist in cells_in_range:
latcell, loncell = cell.split(',')
latcell = float(latcell)
loncell = float(loncell)
actualdist = info.distance(latquery, lonquery, latcell, loncell)
outputFilePath = os.path.join(
mainOutputDir,
querysurf + ',' + cell + ',' + str(actualdist) + ".res")
if copy_top_n_percell > 0:
outputDir = os.path.join(
mainOutputDir, querysurf + ',' + cell + ',' + str(actualdist))
copy_topn_results(os.path.join(dbdir, cell), outputDir,
outputFilePath, 4)
# combined = combine_until_dup(outputFilePaths, 1000)
combined = combine_topn_votes(outputFilePaths, float('inf'))
# combined = filter_in_range(combined, querysurf)
# write_scores(querysurf, combined, "/media/data/combined")
[g, y, r, b, o] = check_topn_img(querysurf, combined, topnresults)
if copytopmatch:
match = g or y or r or b or o
copy_top_match(querydir,
querysurf.split('surf.npy')[0], combined, match)
return [g, y, r, b, o]
from android import AndroidReader
import numpy as np
from info import distance
f = open('/media/DATAPART2/query5horizontal/gtLatLonNormYaw.txt')
m = {}
for line in f:
data = line.split()
name = data[0]
lat, lon = float(data[1]), float(data[2])
m[name] = (lat, lon)
dss = []
for a in AndroidReader('/media/DATAPART2/query5horizontal'):
if a.name in m:
d = m[a.name]
ds = distance(a.lat, a.lon, d[0], d[1])
ds = float(ds)
dss.append(ds)
dss.append(-ds)
print ds
else:
print 'skipped', a.name, a.lat, a.lon
print 'mean', np.mean(dss)
print 'var', np.var(dss)
def similar(view): return distance(view.lat, view.lon, qlat, qlon) < thresh
def match(C, Q):
if C.shuffle_cells:
C._dbdir = None
if C.override_cells:
INFO('override cells')
cells_in_range = [(c,0) for c in C.override_cells]
else:
# compute closest cells
closest_cells = util.getclosestcells(Q.query_lat, Q.query_lon, C.dbdir)
if C.restrict_cells:
closest_cells = filter(lambda c: c[0] in C.restrict_cells, closest_cells)
cells_in_range = [(cell, dist)
for cell, dist in closest_cells[0:C.ncells]
if dist < C.cellradius + C.ambiguity + C.matchdistance]
INFO('Using %d cells' % len(cells_in_range))
if C.shuffle_cells:
import reader
sr = reader.get_reader('sift')
supercell = sr.get_supercelldir(
C.dbdir,
[c for (c,d) in cells_in_range],
C.overlap_method)
C._dbdir = supercell
if not cells_in_range:
raise LocationOutOfRangeError
# cache for fuzz runs
if C.cacheEnable:
key = derive_key(C, cells_in_range, Q.siftname)
if key in cache:
print 'cache hit'
return cache[key]
else:
print 'cache miss'
# compute output file paths for the cells
cellpath = [c for c,d in cells_in_range]
listofimages = []
if C.one_big_cell:
INFO('Using 1 big cell (%d union)' % len(cells_in_range))
outputFilePaths = [os.path.join(C.matchdir, Q.siftname + ',' + getcellid(cellpath) + ".res")]
#listofimages = lexiconrank.addImagetoList(listofimages, C.dbdir + cellpath)
cellpath = [cellpath]
else:
outputFilePaths = []
for cell, dist in cells_in_range:
if ',' in cell:
latcell, loncell = cell.split(',')
latcell = float(latcell)
loncell = float(loncell)
else:
latcell, loncell = 0,0
actualdist = info.distance(Q.query_lat, Q.query_lon, latcell, loncell)
outputFilePath = os.path.join(C.matchdir, Q.siftname + ',' + cell + ',' + str(actualdist) + ".res")
outputFilePaths.append(outputFilePath)
#listofimages = lexiconrank.addImagetoList(listofimages, C.dbdir + cell)
# start query
query.run_parallel(C, Q, cellpath, outputFilePaths, estimate_threads_avail())
#d, lexiconmatchedimg = lexiconrank.returnTopMatch_random(C.dbdump, listofimages, Q.jpgpath)
# combine results
if C.spatial_comb:
comb_matches = corr.combine_spatial(outputFilePaths)
else:
print outputFilePaths
comb_matches = corr.combine_matches(outputFilePaths)
#geometric consistency reranking
if C.disable_filter_step:
imm = condense2(sorted(comb_matches.iteritems(), key=lambda x: len(x[1]), reverse=True))
rsc_ok = True
else:
imm, rsc_ok = rerank_ransac(comb_matches, C, Q)
if C.weight_by_coverage:
#print 1
ranked = weight_by_coverage(C, Q, imm)
elif C.weight_by_distance:
#print 2
ranked = weight_by_distance(C, Q, imm)
else:
#print 3
ranked = distance_sort(C, Q, imm)
# top 1
stats = check_topn_img(C, Q, ranked, 1)
# return statistics and top result
matchedimg = ranked[0][0]
matches = comb_matches[matchedimg + 'sift.txt']
if C.cacheEnable:
cache[key] = (stats, matchedimg, matches, ranked)
if C.match_callback:
C.match_callback(C, Q, stats, matchedimg, ranked, cells_in_range, rsc_ok)
# compute homography and draw images maybe
if MultiprocessExecution.pool:
MultiprocessExecution.pool.apply_async(compute_hom, [C.pickleable(), Q, ranked, comb_matches])
else:
compute_hom(C, Q, ranked, comb_matches)
### Query Pose Estimation ###
match = any(check_img(C, Q, ranked[0]))
if (C.solve_pose and match and Q.name not in C.pose_remove) or C.pose_param['solve_bad']:
#computePose.draw_dbimage(C, Q, matchedimg, match)
if MultiprocessExecution.pool:
MultiprocessExecution.pool.apply_async(computePose.estimate_pose, [C.pickleable(), Q, matchedimg, match])
else:
computePose.estimate_pose(C, Q, matchedimg, match)
# done
return stats, matchedimg, matches, ranked
def extract_key(match): line = match[0] lat, lon = getlatlonfromdbimagename(C, line) return (match[1], -info.distance(lat, lon, Q.query_lat, Q.query_lon))
def distance3d6(x1, y1, z1, x2, y2, z2):
xydist = distance(x1, y1, x2, y2)
vert = abs(z1 - z2)
return sqrt(xydist**2 + vert**2)
def contribution(target, contributer): lat, lon = getlatlonfromdbimagename(C, target[0]) lat2, lon2 = getlatlonfromdbimagename(C, contributer[0]) dist = info.distance(lat, lon, lat2, lon2) return contrib_function(contributer) * weighting_function(dist)