def show_results(results, names, title=""):
"""
Show the results in a nice grid.
Show only 20 top results
params
------
results: ndarray (., 2)
array containing, in the first column, the id of the result, on
the second, the score of that results
names: ndarray (.)
image database.
"""
fig = plt.figure(title)
for i, result in enumerate(results):
if i == 20:
break
image_name = names[i]
image, _ = load.get_image(image_name)
ax = fig.add_subplot(5, 4, i + 1)
ax.xaxis.set_visible(False)
ax.yaxis.set_visible(False)
if len(image.shape) == 3:
ax.imshow(image)
else:
ax.imshow(image, cmap=cm.gray)
ax.set_title("%02d" % result[1])
ax.imshow(image)
#ax.set_title("%02d" % result[1])
ax.set_title(image_name)
plt.show()
def __init__(self, dest_surface, frames, delay=0.1, xbool=False):
"""
:param dest_surface:
:param frames: список с именами файлов
:param delay: скорость смены кадров
"""
self.image = dest_surface
res = list()
for frame in frames:
image, _, _ = get_image(frame)
if xbool:
image = transform.flip(image, True, False) # зеркалим
res.append(image)
self.frames = res
self.frame_number = 0
self.frame_max = len(frames)
self.start_time = time.time()
self.delay = delay
def main():
pygame.init()
screen = pygame.display.set_mode(DISPLAY)
pygame.display.set_caption('test v0.0.1')
bg, _, _ = get_image('640_winter_lab_by_kooner_cz.png')
entities = pygame.sprite.Group() # Все объекты
hero = Player(0, 50)
entities.add(hero)
platforms, total_level_width, total_level_height = load_level(entities)
camera = Camera(WIN_WIDTH, WIN_HEIGHT, total_level_width,
total_level_height)
timer = pygame.time.Clock()
keys = {
K_UP: False,
K_DOWN: False,
K_LEFT: False,
K_RIGHT: False,
K_SPACE: False,
}
while True:
timer.tick(60)
for e in pygame.event.get():
if e.type == QUIT:
return
if e.type == KEYDOWN:
if e.key == K_ESCAPE:
return
keys[e.key] = True
if e.type == KEYUP:
keys[e.key] = False
hero.update(keys, platforms) # передвижение
camera.update(hero) # центризируем камеру относительно персонажа
screen.blit(bg,
(0, 0)) # каждую итерацию необходимо всё перерисовывать
for e in entities:
screen.blit(e.image, camera.apply(e))
pygame.display.update() # обновление и вывод всех изменений на экран
def show_rate():
current_path = os.path.dirname(__file__)
print('* Current path = ' + current_path)
print('Enter path of dataset')
while True:
try:
data_path = input()
total = len(fnmatch.filter(os.listdir(data_path), '*.png'))
break
except FileNotFoundError:
print('Wrong path. Try again.')
order = choose_process()
count = letter_correct = correct = 0
for f in glob.glob(data_path + '/*.png'):
count += 1
original_image, label = get_image(f)
try:
processed_image = process(original_image, order)
if processed_image.text == label:
correct += 1
for i in range(min(len(processed_image.text), len(label))):
if processed_image.text[i] == label[i]:
letter_correct += 1
except:
print('Error occured in processing image ', label)
print('\r{0: .2f}% complete..'.format((count * 100) / total), end='')
print('')
print('\tOut of {0} letters {1} were correctly read.'
'Success Rate: {2:.2f}%'.format(
total * 6, letter_correct, (letter_correct * 100) / (total * 6)))
print('\tOut of {0} captcha images, {1} were correctly read.'
'Success Rate: {2:.2f}%'.format(total, correct,
(correct * 100) / total))
print('Press any key to continue.')
input()
def score_results(coords, desc, search_results, names, voc, verbose=False):
"""
Scores the 200 best results
params
------
coords:
desc:
search_results:
names: ndarray,
image database
voc: ndarray
vocabulary
verbose: boolean, optional, default: False
Make output more verbose
returns:
search_results: ndarray
indxs, scores
"""
# FIXME doxstring
for j, (result, score) in enumerate(search_results):
if verbose:
print "Scoring %d / %d" % (j, len(search_results))
im2, mask2 = get_image(names[result, 0])
interest_points = get_interest_points(mask2)
desc2, coords2 = compute_boundary_desc(im2,
mask2,
interest_points)
if desc2:
search_results[j, 1] += score_(desc, desc2, coords, coords2)
idxs = search_results[:, 1].argsort()[::-1]
return search_results[idxs, :]
queries_total = 0
max_im = None
for test_query_index, query in enumerate(ground_truth):
if test_query_index % 10 == 0:
print "Computed %d images" % test_query_index
if max_im and test_query_index == max_im:
break
query_file_name = query.keys()[0]
positive_file_names = set(query.values()[0][0])
ignore_file_names = set(query.values()[0][1])
if query_file_name not in all_names:
continue
im, mask = load.get_image(query_file_name)
interest_points = mem.cache(get_interest_points)(mask)
desc, coords = mem.cache(compute_boundary_desc)(im, mask, interest_points)
visual_words = compute_visual_words(desc, voc)
if visual_words is None:
continue
#search_results = retrieval.search2(visual_words, postings, max_im=20)
query_document, _ = np.histogram(visual_words,
bins=np.arange(len(voc) + 1))
search_results = retrieval2.search(query_document, max_im=20)
#search_results2 = mem.cache(search2)(visual_words,postings,max_im=20)
indices = search_results[:, 0].astype(int)
search_file_names = all_names[indices]
queries_total += 1
def __init__(self, x, y):
sprite.Sprite.__init__(self)
self.image, w, h = get_image('platform.png')
self.rect = Rect(x, y, w, h)
queries_total = 0
max_im = None
for test_query_index, query in enumerate(ground_truth):
if test_query_index % 10 == 0:
print "Computed %d images" % test_query_index
if max_im and test_query_index == max_im:
break
query_file_name = query.keys()[0]
positive_file_names = set(query.values()[0][0])
ignore_file_names = set(query.values()[0][1])
if query_file_name not in all_names:
continue
im, mask = load.get_image(query_file_name)
interest_points = mem.cache(get_interest_points)(mask)
desc, coords = mem.cache(compute_boundary_desc)(im, mask, interest_points)
visual_words = compute_visual_words(desc, voc)
if visual_words is None:
continue
#search_results = retrieval.search2(visual_words, postings, max_im=20)
query_document, _ = np.histogram(visual_words,
bins=np.arange(len(voc) + 1))
search_results = retrieval2.search(query_document, max_im=20)
#search_results2 = mem.cache(search2)(visual_words,postings,max_im=20)
indices = search_results[:, 0].astype(int)
search_file_names = all_names[indices]
queries_total += 1
import numpy as np
from sklearn.externals.joblib import Memory
from load import get_image
from descriptors import get_interest_points, compute_boundary_desc
from histograms import compute_visual_words
from ransac_homography import ransac
from retrieval import match_descriptors
from draw import show_matched_desc
mem = Memory(cachedir='.')
# FIXME Choose a couple of images, and load them properly
vocabulary = np.load('./data/vocabulary.npy')
im1, mask1 = get_image('Henry_Moore_Cut_0006.jpg', test=True)
im2, mask2 = get_image('Henry_Moore_Cut_0034.jpg', test=True)
interest_points = mem.cache(get_interest_points)(mask1)
desc1, coords1 = mem.cache(compute_boundary_desc)(im1,
mask1,
interest_points)
voc1 = vocabulary[compute_visual_words(desc1, vocabulary)]
interest_points = mem.cache(get_interest_points)(mask2)
desc2, coords2 = mem.cache(compute_boundary_desc)(im2,
mask2,
interest_points)
voc2 = vocabulary[compute_visual_words(desc2, vocabulary)]
# Use, as for a sift matching a nearest neighbour / second nearest neighbour
# matching.
