def recognise(image, addr, extras):
result = ""
x = image.width - 1
channels = getChannels(image)
bestBounds = []
#cv.NamedWindow("pic", 1)
#cv.NamedWindow("cols", 0)
while len(result) < nSegs and x >= minW:
x = cap.getBound(image, cap.CAP_BOUND_RIGHT, start=x)
ratings = []
for w in xrange(minW, min(maxW + 1, x)):
bounds = findBounds(image, x, w)
subImage = cap.getSubImage(image, bounds)
flags = findColors(subImage)
for index, flag in enumerate(flags):
if not flag: continue
seg = getSegment(channels[index], image, bounds)
seg = cap.flattenImage(adjustSize(seg, segSize))
guesses = ann.run(seg)
charIndex = cap.argmax(guesses)
ratings.append((guesses[charIndex], charIndex, index, bounds, seg))
best = max(ratings, key=itemgetter(0))
result += charset[best[1]]
bestChannel = channels[best[2]]
cv.SetImageROI(bestChannel, best[3])
cv.Set(bestChannel, 96, bestChannel)
cv.ResetImageROI(bestChannel)
bestBounds.append(best[3])
bestW = best[3][2]
x -= bestW
#print ann.run(best[4])
cap.processExtras([cap.drawComponents(image, bestBounds)], addr, extras, cap.CAP_STAGE_RECOGNISE)
return result[::-1]
def recogniseOne(seg):
resized = adjustSize(seg, segSize)
array = cap.flattenImage(resized)
guesses = ann.run(array)
ichar = cap.argmax(guesses)
guess = guesses[ichar]
return (guess, ichar)
def recogniseOne(seg):
resized = adjustSize(seg, segSize)
array = cap.flattenImage(resized)
guesses = ann.run(array)
ichar = cap.argmax(guesses)
guess = guesses[ichar]
return (guess, ichar)
def recognise(image, addr, extras):
result = ""
x = image.width - 1
channels = getChannels(image)
bestBounds = []
#cv.NamedWindow("pic", 1)
#cv.NamedWindow("cols", 0)
while len(result) < nSegs and x >= minW:
x = cap.getBound(image, cap.CAP_BOUND_RIGHT, start=x)
ratings = []
for w in xrange(minW, min(maxW + 1, x)):
bounds = findBounds(image, x, w)
subImage = cap.getSubImage(image, bounds)
flags = findColors(subImage)
for index, flag in enumerate(flags):
if not flag: continue
seg = getSegment(channels[index], image, bounds)
seg = cap.flattenImage(adjustSize(seg, segSize))
guesses = ann.run(seg)
charIndex = cap.argmax(guesses)
ratings.append(
(guesses[charIndex], charIndex, index, bounds, seg))
best = max(ratings, key=itemgetter(0))
result += charset[best[1]]
bestChannel = channels[best[2]]
cv.SetImageROI(bestChannel, best[3])
cv.Set(bestChannel, 96, bestChannel)
cv.ResetImageROI(bestChannel)
bestBounds.append(best[3])
bestW = best[3][2]
x -= bestW
#print ann.run(best[4])
cap.processExtras([cap.drawComponents(image, bestBounds)], addr, extras,
cap.CAP_STAGE_RECOGNISE)
return result[::-1]
def mainLoop():
n_iter = 0
last_save = 0
min_test_MSE = 1.0
max_iters_after_save = 50
try:
while True:
n_iter += 1
print "Iteration: %5d " % (n_iter),
seg_copy = map(lambda (c, seg): (c, cv.CloneImage(seg)), segments)
seg_copy = map(lambda (c, seg): (c, spoil(seg)), seg_copy)
shuffle(seg_copy)
f = open(train_file, "w")
f.write("%d %d %d\n" % (len(segments), num_input, num_output))
for c, image in seg_copy:
image = adjustSize(image, (segW, segH))
for y in range(image.height):
for x in range(image.width):
n = image[y, x] / 159.375 - 0.8
f.write("%f " % n)
f.write("\n")
n = charset.index(c)
f.write("-1 " * n + "1" + " -1" * (num_output - n - 1) + "\n")
f.close()
train = libfann.training_data()
train.read_train_from_file(train_file)
ann.train_epoch(train)
train.destroy_train()
print "Train MSE: %f " % (ann.get_MSE()),
print "Train bit fail: %5d " % (ann.get_bit_fail()),
ann.test_data(test)
mse = ann.get_MSE()
print "Test MSE: %f " % (mse),
print "Test bit fail: %5d " % (ann.get_bit_fail()),
if mse < min_test_MSE:
min_test_MSE = mse
ann.save(ann_file)
last_save = n_iter
print "saved",
if n_iter - last_save > max_iters_after_save: break
print
except KeyboardInterrupt:
print "Interrupted by user."
def mainLoop():
n_iter = 0
last_save = 0
min_test_MSE = 1.0
max_iters_after_save = 50
try:
while True:
n_iter += 1
print "Iteration: %5d " % (n_iter),
seg_copy = map(lambda (c, seg): (c, cv.CloneImage(seg)), segments)
seg_copy = map(lambda (c, seg): (c, spoil(seg)), seg_copy)
shuffle(seg_copy)
f = open(train_file, "w")
f.write("%d %d %d\n" % (len(segments), num_input, num_output))
for c, image in seg_copy:
image = adjustSize(image, (segW, segH))
for y in range(image.height):
for x in range(image.width):
n = image[y, x] / 159.375 - 0.8
f.write("%f " % n)
f.write("\n")
n = charset.index(c)
f.write("-1 " * n + "1" + " -1" * (num_output - n - 1) + "\n")
f.close()
train = libfann.training_data()
train.read_train_from_file(train_file)
ann.train_epoch(train)
train.destroy_train()
print "Train MSE: %f " % (ann.get_MSE()),
print "Train bit fail: %5d " % (ann.get_bit_fail()),
ann.test_data(test)
mse = ann.get_MSE()
print "Test MSE: %f " % (mse),
print "Test bit fail: %5d " % (ann.get_bit_fail()),
if mse < min_test_MSE:
min_test_MSE = mse
ann.save(ann_file)
last_save = n_iter
print "saved",
if n_iter - last_save > max_iters_after_save: break
print
except KeyboardInterrupt: print "Interrupted by user."
from pyfann import libfann
from lconsts import segW, segH, num_input, num_output, charset, train_file, ann_file
sys.path.append("..")
from cap import resizeFit as adjustSize, segment_dir
segments = os.listdir(segment_dir)
f = open(train_file, "w")
f.write("%d %d %d\n" % (len(segments), num_input, num_output))
for name in segments:
image = cv.LoadImage(os.path.join(segment_dir, name), cv.CV_LOAD_IMAGE_GRAYSCALE)
image = adjustSize(image, (segW, segH))
for y in range(image.height):
for x in range(image.width):
n = image[y, x] / 159.375 - 0.8
f.write("%f " % n)
f.write("\n")
c = os.path.splitext(name)[0][0]
n = charset.index(c)
f.write("-1 " * n + "1" + " -1" * (num_output - n - 1) + "\n")
f.close()
print "Samples: %d" % len(segments)
print "Input: %d" % num_input
print "Output: %d" % num_output
from pyfann import libfann
from lconsts import segW, segH, num_input, num_output, charset, train_file, ann_file
sys.path.append("..")
from cap import resizeFit as adjustSize, segment_dir
segments = os.listdir(segment_dir)
f = open(train_file, "w")
f.write("%d %d %d\n" % (len(segments), num_input, num_output))
for name in segments:
image = cv.LoadImage(os.path.join(segment_dir, name),
cv.CV_LOAD_IMAGE_GRAYSCALE)
image = adjustSize(image, (segW, segH))
for y in range(image.height):
for x in range(image.width):
n = image[y, x] / 159.375 - 0.8
f.write("%f " % n)
f.write("\n")
c = os.path.splitext(name)[0][0]
n = charset.index(c)
f.write("-1 " * n + "1" + " -1" * (num_output - n - 1) + "\n")
f.close()
print "Samples: %d" % len(segments)
print "Input: %d" % num_input
print "Output: %d" % num_output
