%matplotlib inline
from utils import *
[*] Test mode : True food1 : 620 food2 : 728 food3 : 353 train : 1139 test : 562
<matplotlib.figure.Figure at 0x7f090c83e290>
train_fs = build_feature_vecture(train)
test_fs = build_feature_vecture(test)
save_pickle(train_fs, 'train_fa.pickle')
0% 100% [####################] | ETA[sec]: 0.000 Total time elapsed: 0.774 sec 0% 100% [####################] | ETA[sec]: 0.000 Total time elapsed: 0.782 sec
There exists train_fa.pickle ============================= last modified: Mon Nov 17 16:59:46 2014 size: 3.0MiB ============================= Will you overwrite it? [Y/n] : n
from sklearn import datasets, svm, metrics, cross_validation
print "Start learning"
start_time = time.time()
clf = svm.SVC(gamma=0.001)
clf.fit(train_fs, train_labels[:len(train_fs)])
end_time = time.time()
print "Finished learning"
print("Elapsed time was %g seconds" % (end_time - start_time))
Start learning Finished learning Elapsed time was 0.0159569 seconds
fname = get_image_by_class(foods,1)
plt.figure(1)
plt.imshow(imread(fname))
classify(clf, fname)
result : 36 -> 1 0.0343919 seconds
img = imread(foods[4][0])
print foods[4][1]
print img.shape
#fig, axs = plt.subplots(1,3)
plt.figure(1)
plt.imshow(img)
plt.colorbar()
plt.figure(2)
plt.imshow(color.rgb2hsv(img))
plt.colorbar()
orientations = 9
cell = 3
block = 1
f, hog = feature.hog(color.rgb2gray(img),
orientations=orientations,
pixels_per_cell=(cell, cell),
cells_per_block=(block, block),
visualise =True)
plt.figure(3)
plt.imshow(hog)
plt.colorbar()
10 (188, 220, 3)
<matplotlib.colorbar.Colorbar instance at 0x7f80939dd8c0>
from skimage.util import img_as_float
from skimage import exposure
from skimage.morphology import disk
from skimage.filter import rank
def plot_img_and_hist(img, axes, bins=256):
"""Plot an image along with its histogram and cumulative histogram.
"""
img = img_as_float(img)
ax_img, ax_hist = axes
ax_cdf = ax_hist.twinx()
# Display image
ax_img.imshow(img, cmap=plt.cm.gray)
#ax_img.set_axis_off()
# Display histogram
ax_hist.hist(img.ravel(), bins=bins, histtype='step', color='black')
ax_hist.ticklabel_format(axis='y', style='scientific', scilimits=(0, 0))
ax_hist.set_xlabel('Pixel intensity')
ax_hist.set_xlim(0, 1)
ax_hist.set_yticks([])
# Display cumulative distribution
img_cdf, bins = exposure.cumulative_distribution(img, bins)
ax_cdf.plot(bins, img_cdf, 'r')
ax_cdf.set_yticks([])
return ax_img, ax_hist, ax_cdf
plt.rcParams['font.size'] = 9
fig, axes = plt.subplots(2, 3, figsize=(16,8))
ax_img, ax_hist, ax_cdf = plot_img_and_hist(img, axes[:, 0])
ax_img.set_title('Low contrast image')
ax_hist.set_ylabel('Number of pixels')
fig.subplots_adjust(wspace=0.4)
img = imread(foods[4][0])
print foods[4][1]
print img.shape
f, hog = feature.hog(color.rgb2gray(img),
orientations=orientations,
pixels_per_cell=(cell, cell),
cells_per_block=(block, block),
visualise =True)
plt.figure(1)
plt.imshow(hog)
plt.colorbar()