def _select_ith_kpt(fx):
logger.info('[interact] viewing ith=%r keypoint' % fx)
# Get the fx-th keypiont
kp, sift = kpts[fx], vecs[fx]
# Draw the image with keypoint fx highlighted
_viz_keypoints(fnum, (2, 1, 1), sel_fx=fx,
**kwargs) # MAYBE: remove kwargs
# Draw the selected feature
nRows, nCols, px = (2, 3, 3)
draw_feat_row(chip, fx, kp, sift, fnum, nRows, nCols, px, None)
def _select_ith_kpt(self, fx):
logger.info('[interact] viewing ith=%r keypoint' % fx)
# Get the fx-th keypiont
kp, sift = self.kpts[fx], self.vecs[fx]
# Draw the image with keypoint fx highlighted
self.plot(self.fnum, (2, 1, 1), sel_fx=fx)
# Draw the selected feature
nRows, nCols, px = (2, 3, 3)
draw_feat_row(self.chip, fx, kp, sift, self.fnum, nRows, nCols, px,
None)
def _select_fxth_kpt(fx):
from wbia.plottool.viz_featrow import draw_feat_row
# Get the fx-th keypiont
chip = ibs.get_annot_chips(aid, config2_=config2_)
kp = ibs.get_annot_kpts(aid, config2_=config2_)[fx]
sift = ibs.get_annot_vecs(aid, config2_=config2_)[fx]
# Draw chip + keypoints + highlighted plots
_chip_view(pnum=(2, 1, 1), sel_fx=fx)
# ishow_chip(ibs, aid, fnum=None, fx=fx, config2_=config2_, **kwargs)
# Draw the selected feature plots
nRows, nCols, px = (2, 3, 3)
draw_feat_row(chip, fx, kp, sift, fnum, nRows, nCols, px, None)
def show_nearest_descriptors(ibs,
qaid,
qfx,
fnum=None,
stride=5,
qreq_=None,
**kwargs):
r"""
Args:
ibs (wbia.IBEISController): image analysis api
qaid (int): query annotation id
qfx (int): query feature index
fnum (int): figure number
stride (int):
consecutive_distance_compare (bool):
CommandLine:
# Find a good match to inspect
python -m wbia.viz.interact.interact_matches --test-testdata_match_interact --show --db PZ_MTEST --qaid 3
# Now inspect it
python -m wbia.viz.viz_nearest_descriptors --test-show_nearest_descriptors --show --db PZ_MTEST --qaid 3 --qfx 879
python -m wbia.viz.viz_nearest_descriptors --test-show_nearest_descriptors --show
python -m wbia.viz.viz_nearest_descriptors --test-show_nearest_descriptors --db PZ_MTEST --qaid 3 --qfx 879 --diskshow --save foo.png --dpi=256
SeeAlso:
plottool.viz_featrow
~/code/plottool/plottool/viz_featrow.py
Example:
>>> # DISABLE_DOCTEST
>>> from wbia.viz.viz_nearest_descriptors import * # NOQA
>>> import wbia
>>> # build test data
>>> if True:
>>> import matplotlib as mpl
>>> from wbia.scripts.thesis import TMP_RC
>>> mpl.rcParams.update(TMP_RC)
>>> qreq_ = wbia.testdata_qreq_()
>>> ibs = wbia.opendb('PZ_MTEST')
>>> qaid = qreq_.qaids[0]
>>> qfx = ut.get_argval('--qfx', type_=None, default=879)
>>> fnum = None
>>> stride = 5
>>> # execute function
>>> skip = False
>>> result = show_nearest_descriptors(ibs, qaid, qfx, fnum, stride,
>>> draw_chip=True,
>>> draw_warped=True,
>>> draw_unwarped=False,
>>> draw_desc=False, qreq_=qreq_)
>>> # verify results
>>> print(result)
>>> pt.show_if_requested()
"""
import wbia.plottool as pt # NOQA
consecutive_distance_compare = True
draw_chip = kwargs.get('draw_chip', False)
draw_desc = kwargs.get('draw_desc', True)
draw_warped = kwargs.get('draw_warped', True)
draw_unwarped = kwargs.get('draw_unwarped', True)
# skip = kwargs.get('skip', True)
# Plots the nearest neighbors of a given feature (qaid, qfx)
if fnum is None:
fnum = df2.next_fnum()
try:
# Flann NN query
(qfx, qfx2_daid, qfx2_dfx, qfx2_dist, K,
Knorm) = get_annotfeat_nn_index(ibs, qaid, qfx, qreq_=qreq_)
# Adds metadata to a feature match
def get_extract_tuple(aid, fx, k=-1):
rchip = ibs.get_annot_chips(aid)
kp = ibs.get_annot_kpts(aid)[fx]
sift = ibs.get_annot_vecs(aid)[fx]
if not ut.get_argflag('--texknormplot'):
aidstr = vh.get_aidstrs(aid)
nidstr = vh.get_nidstrs(ibs.get_annot_nids(aid))
id_str = ' ' + aidstr + ' ' + nidstr + ' fx=%r' % (fx, )
else:
id_str = nidstr = aidstr = ''
info = ''
if k == -1:
if pt.is_texmode():
info = '\\vspace{1cm}'
info += 'Query $\\mathbf{d}_i$'
info += '\n\\_'
info += '\n\\_'
else:
if len(id_str) > '':
info = 'Query: %s' % (id_str, )
else:
info = 'Query'
type_ = 'Query'
elif k < K:
type_ = 'Match'
if ut.get_argflag('--texknormplot') and pt.is_texmode():
# info = 'Match:\n$k=%r$, $\\frac{||\\mathbf{d}_i - \\mathbf{d}_j||}{Z}=%.3f$' % (k, qfx2_dist[0, k])
info = '\\vspace{1cm}'
info += 'Match: $\\mathbf{d}_{j_%r}$\n$\\textrm{dist}=%.3f$' % (
k,
qfx2_dist[0, k],
)
# info += '\n$s_{\\tt{LNBNN}}=%.3f$' % (qfx2_dist[0, K + Knorm - 1] - qfx2_dist[0, k])
info += '\n$s=%.3f$' % (qfx2_dist[0, K + Knorm - 1] -
qfx2_dist[0, k])
else:
info = 'Match:%s\nk=%r, dist=%.3f' % (id_str, k,
qfx2_dist[0, k])
info += '\nLNBNN=%.3f' % (qfx2_dist[0, K + Knorm - 1] -
qfx2_dist[0, k])
elif k < Knorm + K:
type_ = 'Norm'
if ut.get_argflag('--texknormplot') and pt.is_texmode():
# info = 'Norm: $j_%r$\ndist=%.3f' % (id_str, k, qfx2_dist[0, k])
info = '\\vspace{1cm}'
info += 'Norm: $j_%r$\n$\\textrm{dist}=%.3f$' % (
k, qfx2_dist[0, k])
info += '\n\\_'
else:
info = 'Norm: %s\n$k=%r$, dist=$%.3f$' % (
id_str,
k,
qfx2_dist[0, k],
)
else:
raise Exception('[viz] problem k=%r')
return (rchip, kp, sift, fx, aid, info, type_)
extracted_list = []
# Remember the query sift feature
extracted_list.append(get_extract_tuple(qaid, qfx, -1))
origsift = extracted_list[0][2]
skipped = 0
for k in range(K + Knorm):
# if qfx2_daid[0, k] == qaid and qfx2_dfx[0, k] == qfx:
if qfx2_daid[0, k] == qaid:
skipped += 1
continue
tup = get_extract_tuple(qfx2_daid[0, k], qfx2_dfx[0, k], k)
extracted_list.append(tup)
# Draw the _select_ith_match plot
nRows = len(extracted_list)
if stride is None:
stride = nRows
# Draw selected feature matches
prevsift = None
px = 0 # plot offset
px_shift = 0 # plot stride shift
nExtracted = len(extracted_list)
featrow_kw = dict(
draw_chip=draw_chip,
draw_desc=draw_desc,
draw_warped=draw_warped,
draw_unwarped=draw_unwarped,
)
if ut.get_argflag('--texknormplot'):
featrow_kw['ell_color'] = pt.ORANGE
featrow_kw['ell_linewidth'] = 1
featrow_kw['arm1_lw'] = 0.5
featrow_kw['stroke'] = 0
pass
for listx, tup in enumerate(extracted_list):
(rchip, kp, sift, fx, aid, info, type_) = tup
if listx % stride == 0:
# Create a temporary nRows and fnum in case we are splitting
# up nearest neighbors into separate figures with stride
_fnum = fnum + listx
_nRows = min(nExtracted - listx, stride)
px_shift = px
df2.figure(fnum=_fnum, docla=True, doclf=True)
px_ = px - px_shift
px = draw_feat_row(
rchip,
fx,
kp,
sift,
_fnum,
_nRows,
px=px_,
prevsift=prevsift,
origsift=origsift,
aid=aid,
info=info,
type_=type_,
**featrow_kw,
)
px += px_shift
if prevsift is None or consecutive_distance_compare:
prevsift = sift
# df2.adjust_subplots(hspace=.85, wspace=0, top=.95, bottom=.087, left=.05, right=.95)
except Exception as ex:
logger.info('[viz] Error in show nearest descriptors')
logger.info(ex)
raise
def select_ith_match(self, mx):
"""
Selects the ith match and visualizes and prints information concerning
features weights, keypoint details, and sift descriptions
"""
import wbia.plottool as pt
from wbia.plottool import viz_featrow
from wbia.plottool import interact_helpers as ih
fnum = self.fnum
same_fig = self.same_fig
rchip1 = self.rchip1
rchip2 = self.rchip2
self.mx = mx
print('+--- SELECT --- ')
print('... selecting mx-th=%r feature match' % mx)
fsv = self.fsv
fs = self.fs
print('score stats:')
print(ut.repr2(ut.get_stats(fsv, axis=0), nl=1))
print('fsv[mx] = %r' % (fsv[mx], ))
print('fs[mx] = %r' % (fs[mx], ))
# ----------------------
# Get info for the select_ith_match plot
self.mode = 1
# Get the mx-th feature match
fx1, fx2 = self.fm[mx]
# Older info
fscore2 = self.fs[mx]
fk2 = None if self.fk is None else self.fk[mx]
kp1, kp2 = self.kpts1[fx1], self.kpts2[fx2]
vecs1, vecs2 = self.vecs1[fx1], self.vecs2[fx2]
info1 = '\nquery'
info2 = '\nk=%r fscore=%r' % (fk2, fscore2)
# self.last_fx = fx1
self.last_fx = fx1
# Extracted keypoints to draw
extracted_list = [
(rchip1, kp1, vecs1, fx1, 'aid1', info1),
(rchip2, kp2, vecs2, fx2, 'aid2', info2),
]
# Normalizng Keypoint
# if hasattr(cm, 'filt2_meta') and 'lnbnn' in cm.filt2_meta:
# qfx2_norm = cm.filt2_meta['lnbnn']
# # Normalizing chip and feature
# (aid3, fx3, normk) = qfx2_norm[fx1]
# rchip3 = ibs.get_annot_chips(aid3)
# kp3 = ibs.get_annot_kpts(aid3)[fx3]
# sift3 = ibs.get_annot_vecs(aid3)[fx3]
# info3 = '\nnorm %s k=%r' % (vh.get_aidstrs(aid3), normk)
# extracted_list.append((rchip3, kp3, sift3, fx3, aid3, info3))
# else:
# pass
# #print('WARNING: meta doesnt exist')
# ----------------------
# Draw the select_ith_match plot
nRows, nCols = len(extracted_list) + same_fig, 3
# Draw matching chips and features
sel_fm = np.array([(fx1, fx2)])
pnum1 = (nRows, 1, 1) if same_fig else (1, 1, 1)
vert = self.vert if self.vert is not None else False
self.chipmatch_view(
pnum=pnum1,
ell_alpha=0.4,
ell_linewidth=1.8,
colors=pt.BLUE,
sel_fm=sel_fm,
vert=vert,
)
# Draw selected feature matches
px = nCols * same_fig # plot offset
prevsift = None
if not same_fig:
# fnum2 = fnum + len(viz.FNUMS)
fnum2 = self.fnum2
fig2 = pt.figure(fnum=fnum2, docla=True, doclf=True)
else:
fnum2 = fnum
for (rchip, kp, sift, fx, aid, info) in extracted_list:
px = viz_featrow.draw_feat_row(
rchip,
fx,
kp,
sift,
fnum2,
nRows,
nCols,
px,
prevsift=prevsift,
aid=aid,
info=info,
)
prevsift = sift
if not same_fig:
ih.connect_callback(fig2, 'button_press_event', self.on_click)