def run():
flag = True # A bool to control the running of the program
hide = True # A bool to control superimposition of image
view = 1 # which view to be change with given commands, 1-Front, 2-Side, 3-Top
with open('lengths.p', 'rb') as fp: # to load the template from the PC
lens = pickle.load(fp)
img = outline('man_2.jpg') # outline is captured
simg = outline('side.jpg')
while flag:
if hide:
front = np.zeros((500, 480))
side = np.zeros((500, 480))
else:
front = img.copy()
side = simg.copy()
top = np.zeros((480, 480))
display(front, side, top, lens) # to show the templates
flag, hide, view = controller(lens, flag, hide,
view) # A user interface to get inputs
def color_segment(self,text_color=RGBPixel(255,255,0),\
instaff_color=RGBPixel(0,255,0),\
other_color=RGBPixel(100,100,100),
classified_color=None,
classified_box=False):
""" Segment image in three with colors
Segment the image into three parts:
- text
- inside staff
- Others/relevant for Classifier
Keyword arguments:
text_color --- What color to use for text ccs
instaff_color --- The color to use for in-staff ccs
other_color --- The color of the rest.
classified_color --- If set we will try to classify stuff in the
image and give them the given color
classified_box --- If set we will try to classify and but instead of
highlight I will box them.
"""
ret = self.to_rgb().to_onebit().to_rgb()
classify = False
if not(classified_color is None and classified_box is None):
classify = True
if classified_color is None:
classified_color = RGBPixel(255,0,0)
text,instaff,other,classified = self.segment(classify=classify)
# Painting inside staff things green
for c in instaff:
ret.highlight(c,instaff_color)
# Painting relevant ccs' red.
for c in other:
ret.highlight(c,other_color)
for c in classified:
if classified_box:
outline(ret,c,width=2.0,color=classified_color)
else:
ret.highlight(c,classified_color)
# Painting text yellow
for c in text:
ret.highlight(c,text_color)
return ret
def color_segment(self,text_color=RGBPixel(255,255,0),\
instaff_color=RGBPixel(0,255,0),\
other_color=RGBPixel(100,100,100),
classified_color=None,
classified_box=False):
""" Segment image in three with colors
Segment the image into three parts:
- text
- inside staff
- Others/relevant for Classifier
Keyword arguments:
text_color --- What color to use for text ccs
instaff_color --- The color to use for in-staff ccs
other_color --- The color of the rest.
classified_color --- If set we will try to classify stuff in the
image and give them the given color
classified_box --- If set we will try to classify and but instead of
highlight I will box them.
"""
ret = self.to_rgb().to_onebit().to_rgb()
classify = False
if not (classified_color is None and classified_box is None):
classify = True
if classified_color is None:
classified_color = RGBPixel(255, 0, 0)
text, instaff, other, classified = self.segment(classify=classify)
# Painting inside staff things green
for c in instaff:
ret.highlight(c, instaff_color)
# Painting relevant ccs' red.
for c in other:
ret.highlight(c, other_color)
for c in classified:
if classified_box:
outline(ret, c, width=2.0, color=classified_color)
else:
ret.highlight(c, classified_color)
# Painting text yellow
for c in text:
ret.highlight(c, text_color)
return ret
def test_outline_contains_elements(self):
story = outline()
for s in story:
if isinstance(s.subject, common.Person):
break
else:
raise AssertionError('Person not exists in outline')
for s in story:
if isinstance(s.subject, common.Stage):
break
else:
raise AssertionError('Stage not exits is outline')
for s in story:
if isinstance(s.subject, common.DayTime):
break
else:
raise AssertionError("DayTime not exits is outline")
self.images.append(ret)
return ret
if __name__ == '__main__':
FORMAT = "%(asctime)-15s %(levelname)s [%(name)s.%(funcName)s-%(lineno)s] %(message)s"
logging.basicConfig(level=logging.DEBUG, format=FORMAT)
start = time.time()
init_gamera()
c = Classifier_with_remove(sys.argv[1], float(sys.argv[2]))
d_t = c.d_t()
logging.debug("Loaded Gamera and classifier in %f seconds",
(time.time() - start))
logging.debug("count_of_training=%d, k=%d, e_fp=%f, d_t=%f",
(len(c.stats), c.k, c.e_fp, d_t))
sys.stdout.flush()
start = time.time()
for imgname in sys.argv[3:]:
m = re.match(r"^(.*)\.[^\.]+$", imgname)
noend = m.group(1)
ci = c.classify_image(imgname)
rgbimg = ci.image.to_rgb()
cg = ci.classified_glyphs(d_t)
[outline(rgbimg, g, 3.0, RGBPixel(255, 0, 0)) for g in cg]
rgbimg.save_PNG("class_%s.png" % noend)
print "Saved class_%s.png: %d glypgs found" % (noend, len(cg))
sys.stdout.flush()
print "Parsed %d images in %f seconds" % (len(
sys.argv[3:]), time.time() - start)
import sys
import re
import time
init_gamera()
progress = 0
amount = len(sys.argv[1:])
elapsed = 0
for i, imgname in enumerate(sys.argv[1:]):
progress = ((i + 1) / float(amount)) * 100
m = re.match(r"^(.*)\.[^\.]+$", imgname)
noend = m.group(1)
image = load_image(imgname)
rgbimg = image.to_rgb()
image = image.to_onebit()
ms = remstaves(image)
ccs = ms.image.cc_analysis()
cond = inout_staff_condition(ms.get_staffpos())
for c in ccs:
if cond(c):
# insid
col = RGBPixel(255, 0, 0)
else:
col = RGBPixel(0, 255, 0)
outline(rgbimg, c, 2.0, col)
savename = "insideoutside_%s.png" % noend
rgbimg.save_PNG(savename)
print "Saved %s" % savename
sys.stdout.flush()
else:
relevant_cc = ccs
ret = Classified_image(self,mi,relevant_cc)
self.images.append(ret)
return ret
if __name__ == '__main__':
FORMAT = "%(asctime)-15s %(levelname)s [%(name)s.%(funcName)s-%(lineno)s] %(message)s"
logging.basicConfig(level=logging.DEBUG,format=FORMAT)
start=time.time()
init_gamera()
c = Classifier_with_remove(sys.argv[1],float(sys.argv[2]))
d_t = c.d_t()
logging.debug("Loaded Gamera and classifier in %f seconds",(time.time()-start))
logging.debug("count_of_training=%d, k=%d, e_fp=%f, d_t=%f",(len(c.stats),c.k,c.e_fp,d_t))
sys.stdout.flush()
start=time.time()
for imgname in sys.argv[3:]:
m = re.match(r"^(.*)\.[^\.]+$",imgname)
noend = m.group(1)
ci = c.classify_image(imgname)
rgbimg = ci.image.to_rgb()
cg = ci.classified_glyphs(d_t)
[outline(rgbimg,g,3.0,RGBPixel(255,0,0)) for g in cg]
rgbimg.save_PNG("class_%s.png"%noend)
print "Saved class_%s.png: %d glypgs found"%(noend,len(cg))
sys.stdout.flush()
print "Parsed %d images in %f seconds"%(len(sys.argv[3:]),time.time()-start)
def test_outline_all_actions(self):
actions = outline()
for a in actions:
self.assertIsInstance(a, common.Act)
def test_e_fp(filename,expected_count=10):
init_gamera()
c = Classifier_with_remove()
c.set_k(1)
c.change_features(["volume64regions"])
ci = c.classify_image(filename)
#files = ["mergedyn2.xml", "mergedyn.xml","only-dynamics.xml",
# "newtrain-dynamic.xml", "preomr.xml"]
files = ["preomr.xml","preomr_edited.xml","preomr_edited_cnn.xml"]
import os.path
# try to match with different trainingsets.
for dynamic in ([ d for d in files if os.path.isfile(d) ]):
ci.load_new_training_data(dynamic)
print "%s - count_of_training=%d, k=%d"%(dynamic,len(c.stats),c.k)
result = {} # Push into buckets based on the count of found glyphs.
csv = {}
sys.stdout.flush()
# Try with different epsilon for false_positives: e_fp
for e_fp in arange(0.01,1.01,0.01):
c.e_fp=e_fp
count = len(ci.classified_glyphs())
# Init bucket.
if not result.has_key(count):
result[count] = []
result[count].append((e_fp,c.d_t()))
csv[e_fp] = count
# Find the best match to the wanted result.
k,res,diff = find_nearest(result,expected_count)
confid = [ (len(v),key,v[0][0],v[0][1]) for key,v in result.iteritems() ]
confid2 = [ (key,len(v)) for key,v in result.iteritems() ]
confid.sort(reverse=True)
confid2.sort()
print "efp,count"
for e_fp,c in sorted(csv.iteritems()):
print "%s,%s"%(e_fp,c)
print
print "count,spansize"
for count,spansize in confid2:
print "%s,%s"%(count,spansize)
return
ret = []
for i in range(0,min(10,len(confid))+1):
ret.append(confid[i])
if not result.has_key(expected_count):
print "Never found the desired amount with %s"%dynamic
print "Found in %d(%d): %s"%(k,diff,[r for r in res])
rgbimg = ci.image.to_rgb()
cg = ci.classified_glyphs(res[0].d_t)
[outline(rgbimg,g,3.0,RGBPixel(255,0,0)) for g in cg]
rgbimg.save_PNG("class_%s_%s.png"%(filename,dynamic))
print
def test_e_fp(filename, expected_count=10):
init_gamera()
c = Classifier_with_remove()
c.set_k(1)
c.change_features(["volume64regions"])
ci = c.classify_image(filename)
#files = ["mergedyn2.xml", "mergedyn.xml","only-dynamics.xml",
# "newtrain-dynamic.xml", "preomr.xml"]
files = ["preomr.xml", "preomr_edited.xml", "preomr_edited_cnn.xml"]
import os.path
# try to match with different trainingsets.
for dynamic in ([d for d in files if os.path.isfile(d)]):
ci.load_new_training_data(dynamic)
print "%s - count_of_training=%d, k=%d" % (dynamic, len(c.stats), c.k)
result = {} # Push into buckets based on the count of found glyphs.
csv = {}
sys.stdout.flush()
# Try with different epsilon for false_positives: e_fp
for e_fp in arange(0.01, 1.01, 0.01):
c.e_fp = e_fp
count = len(ci.classified_glyphs())
# Init bucket.
if not result.has_key(count):
result[count] = []
result[count].append((e_fp, c.d_t()))
csv[e_fp] = count
# Find the best match to the wanted result.
k, res, diff = find_nearest(result, expected_count)
confid = [(len(v), key, v[0][0], v[0][1])
for key, v in result.iteritems()]
confid2 = [(key, len(v)) for key, v in result.iteritems()]
confid.sort(reverse=True)
confid2.sort()
print "efp,count"
for e_fp, c in sorted(csv.iteritems()):
print "%s,%s" % (e_fp, c)
print
print "count,spansize"
for count, spansize in confid2:
print "%s,%s" % (count, spansize)
return
ret = []
for i in range(0, min(10, len(confid)) + 1):
ret.append(confid[i])
if not result.has_key(expected_count):
print "Never found the desired amount with %s" % dynamic
print "Found in %d(%d): %s" % (k, diff, [r for r in res])
rgbimg = ci.image.to_rgb()
cg = ci.classified_glyphs(res[0].d_t)
[outline(rgbimg, g, 3.0, RGBPixel(255, 0, 0)) for g in cg]
rgbimg.save_PNG("class_%s_%s.png" % (filename, dynamic))
print
import sys
import re
import time
init_gamera()
progress=0
amount = len(sys.argv[1:])
elapsed = 0
for i,imgname in enumerate(sys.argv[1:]):
progress = ((i+1)/float(amount))*100
m = re.match(r"^(.*)\.[^\.]+$",imgname)
noend = m.group(1)
image = load_image(imgname)
rgbimg = image.to_rgb()
image = image.to_onebit()
ms = remstaves(image)
ccs = ms.image.cc_analysis()
cond = inout_staff_condition(ms.get_staffpos())
for c in ccs:
if cond(c):
# insid
col = RGBPixel(255,0,0)
else:
col = RGBPixel(0,255,0)
outline(rgbimg,c,2.0,col)
savename = "insideoutside_%s.png"%noend
rgbimg.save_PNG(savename)
print "Saved %s"%savename
sys.stdout.flush()