def worker_to_visual_words(wind):
meta = pickle.load(open('../data/traintest.pkl', 'rb'))
all_imagenames = meta['all_imagenames']
point_method = 'Random'
dictionary = pickle.load(open('dictionary%s.pkl' % point_method, 'rb'))
filterBank = create_filterbank()
for j in range(math.ceil(len(all_imagenames) / num_cores)):
img_ind = j * num_cores + wind
if img_ind < len(all_imagenames):
img_name = all_imagenames[img_ind]
print('converting %d-th image %s to visual words' % (img_ind, img_name))
image = cv.imread('../data/%s' % img_name)
image = cv.cvtColor(image, cv.COLOR_BGR2RGB) # convert the image from bgr to rgb
wordMap = get_visual_words(image, dictionary, filterBank)
pickle.dump(wordMap, open('../data/%s_%s.pkl' % (img_name[:-4], point_method), 'wb'))
def get_dictionary(imgPaths, alpha, K, method):
filterBank = create_filterbank()
pixelResponses = np.zeros((alpha * len(imgPaths), 3 * len(filterBank)))
point_count = 0
for i, path in enumerate(imgPaths):
print('-- processing %d/%d' % (i, len(imgPaths)))
image = cv.imread('../data/%s' % path)
image = cv.cvtColor(
image, cv.COLOR_BGR2RGB
) # convert the image from bgr to rgb, OpenCV use BGR by default
# -----fill in your implementation here --------
filt_response = extract_filter_responses(image, filterBank)
if method == 'Random':
points = get_random_points(image, alpha)
for idx in range(alpha):
y = points[0][0][idx]
x = points[0][1][idx]
for n in range(len(filt_response)):
get_pos_value = filt_response[n][y, x]
pixelResponses[point_count, n] = get_pos_value
point_count = point_count + 1
elif method == 'Harris':
points = get_harris_points(image, alpha, k=0.05)
for y, x in points:
for n in range(len(filt_response)):
get_pos_value = filt_response[n][y, x]
pixelResponses[point_count, n] = get_pos_value
point_count = point_count + 1
# ----------------------------------------------
dictionary = KMeans(n_clusters=K,
random_state=0).fit(pixelResponses).cluster_centers_
return dictionary
import pickle
import numpy as np
from python.createFilterBank import create_filterbank
from python.getImageDistance import getImageDistance
traintest_file = open('../data/traintest.pkl', 'rb')
traintest = pickle.load(traintest_file)
traintest_file.close()
test_imagenames = traintest['test_imagenames']
filterBank = create_filterbank()
with open('dictionaryRandom.pkl', 'rb') as handle:
dict_random = pickle.load(handle)
with open('dictionaryHarris.pkl', 'rb') as handle:
dict_harris = pickle.load(handle)
with open('visionRandom.pkl', 'rb') as handle:
train_random_histset = pickle.load(handle)
with open('visionHarris.pkl', 'rb') as handle:
train_harris_histset = pickle.load(handle)
K = 100
rand_eu_acc = []
rand_chi_acc = []
harris_eu_acc = []
harris_chi_acc = []