Merge branch 'master' of github.com:taddeus/licenseplates

Conflicts:
	docs/report.tex

.gitignore

@@ -15,9 +15,3 @@ images/BBB
 images/Images
 images/Infos
 images/licenseplates
-chars
-learning_set
-test_set
-classifier
-classifier-model
-classifier-characters

docs/verslag.tex → docs/report.tex

@@ -45,10 +45,8 @@ in classifying characters on a license plate.
 In short our program must be able to do the following:
 
 \begin{enumerate}
-    \item Use a perspective transformation to obtain an upfront view of license
-          plate.
-    \item Reduce noise where possible to ensure maximum readability.
     \item Extracting characters using the location points in the xml file.
+    \item Reduce noise where possible to ensure maximum readability.
     \item Transforming a character to a normal form.
     \item Creating a local binary pattern histogram vector.
     \item Matching the found vector with a learning set.
@@ -60,7 +58,7 @@ In short our program must be able to do the following:
 The actual purpose of this project is to check if LBP is capable of recognizing
 license plate characters. We knew the LBP implementation would be pretty
 simple. Thus an advantage had to be its speed compared with other license plate 
-recognition implementations, but the uncertainity of whether we could get some
+recognition implementations, but the uncertainty of whether we could get some
 results made us pick Python. We felt Python would not restrict us as much in 
 assigning tasks to each member of the group. In addition, when using the
 correct modules to handle images, Python can be decent in speed.
@@ -70,48 +68,46 @@ correct modules to handle images, Python can be decent in speed.
 Now we know what our program has to be capable of, we can start with the
 implementations.
 
+\subsection{Extracting a letter}
+
+Rewrite this section once we have implemented this properly.
+%NO LONGER VALID!
+%Because we are already given the locations of the characters, we only need to
+%transform those locations using the same perspective transformation used to
+%create a front facing license plate. The next step is to transform the
+%characters to a normalized manner. The size of the letter W is used as a
+%standard to normalize the width of all the characters, because W is the widest
+%character of the alphabet. We plan to also normalize the height of characters,
+%the best manner for this is still to be determined.
+
+%\begin{enumerate}
+%    \item Crop the image in such a way that the character precisely fits the
+%          image.
+%    \item Scale the image to a standard height.
+%    \item Extend the image on either the left or right side to a certain width.
+%\end{enumerate}
+
+%The resulting image will always have the same size, the character contained
+%will always be of the same height, and the character will always be positioned
+%at either the left of right side of the image.
 
 \subsection{Transformation}
 
 A simple perspective transformation will be sufficient to transform and resize
-the plate to a normalized format. The corner positions of license plates in the
+the characters to a normalized format. The corner positions of characters in the
 dataset are supplied together with the dataset.
 
-\subsection{Extracting a letter}
-
-NO LONGER VALID!
-Because we are already given the locations of the characters, we only need to
-transform those locations using the same perspective transformation used to
-create a front facing license plate. The next step is to transform the
-characters to a normalized manner. The size of the letter W is used as a
-standard to normalize the width of all the characters, because W is the widest
-character of the alphabet. We plan to also normalize the height of characters,
-the best manner for this is still to be determined.
-
-\begin{enumerate}
-    \item Crop the image in such a way that the character precisely fits the
-          image.
-    \item Scale the image to a standard height.
-    \item Extend the image on either the left or right side to a certain width.
-\end{enumerate}
-
-The resulting image will always have the same size, the character contained
-will always be of the same height, and the character will alway be positioned
-at either the left of right side of the image.
-
 \subsection{Reducing noise}
 
 Small amounts of noise will probably be suppressed by usage of a Gaussian
 filter. A real problem occurs in very dirty license plates, where branches and
 dirt over a letter could radically change the local binary pattern. A question
 we can ask ourselves here, is whether we want to concentrate ourselves on these
-exceptional cases. By law, license plates have to be readable. Therefore, we
-will first direct our attention at getting a higher score in the 'regular' test
-set before addressing these cases. Considered the fact that the LBP algorithm
-divides a letter into a lot of cells, there is a good change that a great
-number of cells will still match the learning set, and thus still return the
-correct character as a best match. Therefore, we expect the algorithm to be
-very robust when dealing with noisy images.
+exceptional cases. By law, license plates have to be readable. However, the
+provided dataset showed that this does not means they always are. We will have
+to see how the algorithm performs on these plates, however we have good hopes
+that our method will get a good score on dirty plates, as long as a big enough
+part of the license plate remains readable.
 
 \subsection{Local binary patterns}
 Once we have separate digits and characters, we intent to use Local Binary
@@ -128,9 +124,9 @@ form where the pattern is circular.
 \begin{itemize}
 \item Determine the size of the square where the local patterns are being
 registered. For explanation purposes let the square be 3 x 3. \\
-\item The grayscale value of the middle pixel is used a threshold. Every value
-of the pixel around the middle pixel is evaluated. If it's value is greater
-than the threshold it will be become a one else a zero.
+\item The grayscale value of the middle pixel is used as threshold. Every
+value of the pixel around the middle pixel is evaluated. If it's value is
+greater than the threshold it will be become a one else a zero.
 
 \begin{figure}[h!]
 \center
@@ -176,27 +172,29 @@ order. Starting with dividing the pattern in to cells of size 16.
 result is a feature vector of the image.
 
 \item Feed these vectors to a support vector machine. This will ''learn'' which
+<<<<<<< HEAD
 vector indicates what vector is which character. 
+=======
+vectors indicate what letter. 
+>>>>>>> 2e1b2e8c8db4f802d203791a6f03eeca7d0aff70
 
 \end{itemize}
 
 To our knowledge, LBP has yet not been used in this manner before. Therefore,
 it will be the first thing to implement, to see if it lives up to the
-expectations. When the proof of concept is there, it can be used in the final
+expectations. When the proof of concept is there, it can be used in a final
 program.
 
-Important to note is that due to the normalization of characters before
-applying LBP. Therefore, no further normalization is needed on the histograms.
-
-Given the LBP of a character, a Support Vector Machine can be used to classify
-the character to a character in a learning set. The SVM uses
+Later we will show that taking a histogram over the entire image (basically
+working with just one cell) gives us the best results.
 
 \subsection{Matching the database}
 
 Given the LBP of a character, a Support Vector Machine can be used to classify
-the character to a character in a learning set. The SVM uses the collection of
-histograms of an image as a feature vector.  The SVM can be trained with a
-subsection of the given dataset called the ''Learning set''. Once trained, the
+the character to a character in a learning set. The SVM uses a concatenation
+of each cell in an image as a feature vector (in the case we check the entire
+image no concatenation has to be done of course. The SVM can be trained with a
+subset of the given dataset called the ''Learning set''. Once trained, the
 entire classifier can be saved as a Pickle object\footnote{See
 \url{http://docs.python.org/library/pickle.html}} for later usage.
 
@@ -205,11 +203,11 @@ entire classifier can be saved as a Pickle object\footnote{See
 In this section we will describe our implementations in more detail, explaining
 choices we made.
 
-\subsection{Licenseplate retrieval}
+\subsection{Character retrieval}
 
-In order to retrieve the license plate from the entire image, we need to
+In order to retrieve the characters from the entire image, we need to
 perform a perspective transformation. However, to do this, we need to know the 
-coordinates of the four corners of the licenseplate. For our dataset, this is
+coordinates of the four corners of each character. For our dataset, this is
 stored in XML files. So, the first step is to read these XML files.
 
 \paragraph*{XML reader}
@@ -247,9 +245,9 @@ NormalizedImage. When these actions have been completed for each character the
 license plate is usable in the rest of the code.
 
 \paragraph*{Perspective transformation}
-Once we retrieved the cornerpoints of the license plate, we feed those to a
-module that extracts the (warped) license plate from the original image, and
-creates a new image where the license plate is cut out, and is transformed to a
+Once we retrieved the cornerpoints of the character, we feed those to a
+module that extracts the (warped) character from the original image, and
+creates a new image where the character is cut out, and is transformed to a
 rectangle.
 
 \subsection{Noise reduction}
@@ -277,13 +275,6 @@ the dirt, and the fact that the characters are very black, the shape of the
 characters will still be conserved in the LBP, even if there is dirt
 surrounding the character.
 
-\subsection{Character retrieval}
-
-The retrieval of the character is done the same as the retrieval of the license
-plate, by using a perspective transformation. The location of the characters on
-the license plate is also available in de XML file, so this is parsed from that
-as well.
-
 \subsection{Creating Local Binary Patterns and feature vector}
 Every pixel is a center pixel and it is also a value to evaluate but not at the 
 same time. Every pixel is evaluated as shown in the explanation
@@ -337,9 +328,7 @@ value, and what value we decided on.
 The first parameter to decide on, is the $\sigma$ used in the Gaussian blur. To
 find this parameter, we tested a few values, by checking visually what value
 removed most noise out of the image, while keeping the edges sharp enough to
-work with. By checking in the neighbourhood of the value that performed best,
-we where able to 'zoom in' on what we thought was the best value. It turned out
-that this was $\sigma = ?$.
+work with. It turned out the best value is $\sigma = 0.5$.
 
 \subsection{Parameter \emph{cell size}}
 
@@ -352,8 +341,9 @@ be enough cells to properly describe the different areas of the character, and
 the feature vectors will not have enough elements.\\
 \\
 In order to find this parameter, we used a trial-and-error technique on a few
-basic cell sizes, being ?, 16, ?. We found that the best result was reached by
-using ??.
+cell sizes. During this testing, we discovered that a lot better score was
+reached when we take the histogram over the entire image, so with a single
+cell. therefor, we decided to work without cells.
 
 \subsection{Parameters $\gamma$ \& $c$}
 
@@ -374,7 +364,8 @@ checks for each combination of values what the score is. The combination with
 the highest score is then used as our parameters, and the entire SVM will be
 trained using those parameters.\\
 \\
-We found that the best values for these parameters are $c=?$ and $\gamma =?$.
+We found that the best values for these parameters are $c = ?$ and
+$\gamma = ?$.
 
 \section{Results}
 

src/.gitignore

@@ -0,0 +1,2 @@
+*.dat
+results.txt

src/Character.py

@@ -1,4 +1,4 @@
-from LocalBinaryPatternizer import LocalBinaryPatternizer
+from LocalBinaryPatternizer import LocalBinaryPatternizer as LBP
 
 class Character:
     def __init__(self, value, corners, image, filename=None):
@@ -7,7 +7,14 @@ class Character:
         self.image   = image
         self.filename = filename
 
-    def get_feature_vector(self):
-        pattern = LocalBinaryPatternizer(self.image)
+    def get_single_cell_feature_vector(self):
+        if hasattr(self, 'feature'):
+            return
+
+        self.feature = LBP(self.image).single_cell_features_vector()
+
+    def get_feature_vector(self, cell_size=None):
+        pattern = LBP(self.image) if cell_size == None \
+                  else LBP(self.image, cell_size)
 
         return pattern.create_features_vector()

src/Classifier.py

@@ -1,9 +1,11 @@
 from svmutil import svm_train, svm_problem, svm_parameter, svm_predict, \
-        svm_save_model, svm_load_model
+        svm_save_model, svm_load_model, RBF
 
 
 class Classifier:
-    def __init__(self, c=None, gamma=None, filename=None):
+    def __init__(self, c=None, gamma=None, filename=None, cell_size=12):
+        self.cell_size = cell_size
+
         if filename:
             # If a filename is given, load a model from the given filename
             self.model = svm_load_model(filename)
@@ -11,8 +13,8 @@ class Classifier:
             raise Exception('Please specify both C and gamma.')
         else:
             self.param = svm_parameter()
-            self.param.kernel_type = 2  # Radial kernel type
             self.param.C = c  # Soft margin
+            self.param.kernel_type = RBF  # Radial kernel type
             self.param.gamma = gamma  # Parameter for radial kernel
             self.model = None
 
@@ -28,10 +30,12 @@ class Classifier:
         l = len(learning_set)
 
         for i, char in enumerate(learning_set):
-            print 'Training "%s"  --  %d of %d (%d%% done)' \
+            print 'Found "%s"  --  %d of %d (%d%% done)' \
                   % (char.value, i + 1, l, int(100 * (i + 1) / l))
             classes.append(float(ord(char.value)))
-            features.append(char.get_feature_vector())
+            #features.append(char.get_feature_vector())
+            char.get_single_cell_feature_vector()
+            features.append(char.feature)
 
         problem = svm_problem(classes, features)
         self.model = svm_train(problem, self.param)
@@ -48,9 +52,12 @@ class Classifier:
 
         return float(matches) / len(test_set)
 
-    def classify(self, character):
+    def classify(self, character, true_value=None):
         """Classify a character object, return its value."""
-        predict = lambda x: svm_predict([0], [x], self.model)[0][0]
-        prediction_class = predict(character.get_feature_vector())
+        true_value = 0 if true_value == None else ord(true_value)
+        #x = character.get_feature_vector(self.cell_size)
+        character.get_single_cell_feature_vector()
+        p = svm_predict([true_value], [character.feature], self.model)
+        prediction_class = int(p[0][0])
 
-        return chr(int(prediction_class))
+        return chr(prediction_class)

src/ClassifierTest.py

@@ -1,69 +0,0 @@
-#!/usr/bin/python
-from xml_helper_functions import xml_to_LicensePlate
-from Classifier import Classifier
-from cPickle import dump, load
-
-chars = []
-
-for i in range(9):
-    for j in range(100):
-        try:
-            filename = '%04d/00991_%04d%02d' % (i, i, j)
-            print 'loading file "%s"' % filename
-
-            # is nog steeds een licensePlate object, maar die is nu heel anders :P
-            plate = xml_to_LicensePlate(filename) 
-
-            if hasattr(plate, 'characters'):
-                chars.extend(plate.characters)
-        except:
-            print 'epic fail'
-
-print 'loaded %d chars' % len(chars)
-
-dump(chars, file('chars', 'w+'))
-#----------------------------------------------------------------
-chars = load(file('chars', 'r'))[:500]
-learned = []
-learning_set = []
-test_set = []
-
-for char in chars:
-    if learned.count(char.value) > 12:
-        test_set.append(char)
-    else:
-        learning_set.append(char)
-        learned.append(char.value)
-
-#print 'Learning set:', [c.value for c in learning_set]
-#print 'Test set:', [c.value for c in test_set]
-dump(learning_set, file('learning_set', 'w+'))
-dump(test_set, file('test_set', 'w+'))
-#----------------------------------------------------------------
-learning_set = load(file('learning_set', 'r'))
-
-# Train the classifier with the learning set
-classifier = Classifier(c=30, gamma=1)
-classifier.train(learning_set)
-classifier.save('classifier')
-#----------------------------------------------------------------
-classifier = Classifier(filename='classifier')
-test_set = load(file('test_set', 'r'))
-l = len(test_set)
-matches = 0
-
-for i, char in enumerate(test_set):
-    prediction = classifier.classify(char)
-
-    if char.value == prediction:
-        print ':-----> Successfully recognized "%s"' % char.value,
-        matches += 1
-    else:
-        print ':( Expected character "%s", got "%s"' \
-                % (char.value, prediction),
-
-    print '  --  %d of %d (%d%% done)' % (i + 1, l, int(100 * (i + 1) / l))
-
-print '\n%d matches (%d%%), %d fails' % (matches, \
-        int(100 * matches / len(test_set)), \
-        len(test_set) - matches)

src/GrayscaleImage.py

@@ -18,19 +18,23 @@ class GrayscaleImage:
             self.data = data
 
     def __iter__(self):
-        self.__i_x = -1
-        self.__i_y = 0
-        return self
-
-    def next(self):
-        self.__i_x += 1
-        if self.__i_x  == self.width:
-            self.__i_x = 0
-            self.__i_y += 1
-        if self.__i_y == self.height:
-            raise StopIteration
-
-        return  self.__i_y, self.__i_x, self[self.__i_y, self.__i_x]
+        for y in xrange(self.data.shape[0]):
+            for x in xrange(self.data.shape[1]):
+                yield y, x, self.data[y, x]
+
+        #self.__i_x = -1
+        #self.__i_y = 0
+        #return self
+
+    #def next(self):
+    #    self.__i_x += 1
+    #    if self.__i_x  == self.width:
+    #        self.__i_x = 0
+    #        self.__i_y += 1
+    #    if self.__i_y == self.height:
+    #        raise StopIteration
+
+    #    return  self.__i_y, self.__i_x, self[self.__i_y, self.__i_x]
 
     def __getitem__(self, position):
         return self.data[position]

src/Histogram.py

@@ -16,6 +16,12 @@ class Histogram:
     def get_bin_index(self, number):
         return (number - self.min) / ((self.max - self.min) / len(self.bins))
 
+    def normalize(self):
+        minimum = min(self.bins)
+        self.bins = map(lambda b: b - minimum, self.bins)
+        maximum = float(max(self.bins))
+        self.bins = map(lambda b: b / maximum, self.bins)
+
     def intersect(self, other):
         h1 = self.bins
         h2 = other.bins

src/LocalBinaryPatternizer.py

@@ -6,34 +6,37 @@ class LocalBinaryPatternizer:
     def __init__(self, image, cell_size=16):
         self.cell_size = cell_size
         self.image = image
-        self.setup_histograms()
 
     def setup_histograms(self):
         cells_in_width = int(ceil(self.image.width / float(self.cell_size)))
         cells_in_height = int(ceil(self.image.height / float(self.cell_size)))
-        self.features = []
+        self.histograms = []
+
         for i in xrange(cells_in_height):
-            self.features.append([])
+            self.histograms.append([])
+
             for j in xrange(cells_in_width):
-                self.features[i].append(Histogram(256,0,256))
+                self.histograms[i].append(Histogram(256, 0, 256))
 
-    def create_features_vector(self):
-        ''' Walk around the pixels in clokwise order, shifting 1 bit less
-            at each neighbour starting at 7 in the top-left corner. This gives a
-            8-bit feature number of a pixel'''
-        for y, x, value in self.image:
+    def local_binary_pattern(self, y, x, value):
+        return (self.is_pixel_darker(y - 1, x - 1, value) << 7) \
+             | (self.is_pixel_darker(y - 1, x    , value) << 6) \
+             | (self.is_pixel_darker(y - 1, x + 1, value) << 5) \
+             | (self.is_pixel_darker(y    , x + 1, value) << 4) \
+             | (self.is_pixel_darker(y + 1, x + 1, value) << 3) \
+             | (self.is_pixel_darker(y + 1, x    , value) << 2) \
+             | (self.is_pixel_darker(y + 1, x - 1, value) << 1) \
+             | (self.is_pixel_darker(y    , x - 1, value) << 0)
 
-            pattern = (self.is_pixel_darker(y - 1, x - 1, value) << 7) \
-                    | (self.is_pixel_darker(y - 1, x    , value) << 6) \
-                    | (self.is_pixel_darker(y - 1, x + 1, value) << 5) \
-                    | (self.is_pixel_darker(y    , x + 1, value) << 4) \
-                    | (self.is_pixel_darker(y + 1, x + 1, value) << 3) \
-                    | (self.is_pixel_darker(y + 1, x    , value) << 2) \
-                    | (self.is_pixel_darker(y + 1, x - 1, value) << 1) \
-                    | (self.is_pixel_darker(y    , x - 1, value) << 0)
+    def create_features_vector(self):
+        '''Walk around the pixels in clokwise order, shifting 1 bit less at
+        each neighbour starting at 7 in the top-left corner. This gives a 8-bit
+        feature number of a pixel'''
+        self.setup_histograms()
 
+        for y, x, value in self.image:
             cy, cx = self.get_cell_index(y, x)
-            self.features[cy][cx].add(pattern)
+            self.histograms[cy][cx].add(self.local_binary_pattern(y, x, value))
 
         return self.get_features_as_array()
 
@@ -44,4 +47,27 @@ class LocalBinaryPatternizer:
         return (y / self.cell_size, x / self.cell_size)
 
     def get_features_as_array(self):
-        return [h.bins for h in [h for sub in self.features for h in sub]][0]
+        f = []
+
+        # Concatenate all histogram bins
+        for row in self.histograms:
+            for hist in row:
+                f.extend(hist.bins)
+
+        return f
+        #return [h.bins for h in [h for sub in self.histograms for h in sub]][0]
+
+    def get_single_histogram(self):
+        """Create a single histogram of the local binary patterns in the
+        image."""
+        h = Histogram(256, 0, 256)
+
+        for y, x, value in self.image:
+            h.add(self.local_binary_pattern(y, x, value))
+
+        h.normalize()
+
+        return h
+
+    def single_cell_features_vector(self):
+        return self.get_single_histogram().bins

src/NormalizedCharacterImage.py

@@ -5,7 +5,8 @@ from GaussianFilter import GaussianFilter
 
 class NormalizedCharacterImage(GrayscaleImage):
 
-    def __init__(self, image=None, data=None, size=(60, 40), blur=1.1, crop_threshold=0.9):
+    def __init__(self, image=None, data=None, size=(60, 40), blur=1.1, \
+            crop_threshold=0.9):
         if image != None:
             GrayscaleImage.__init__(self, data=deepcopy(image.data))
         elif data != None:
@@ -13,18 +14,17 @@ class NormalizedCharacterImage(GrayscaleImage):
         self.blur = blur
         self.crop_threshold = crop_threshold
         self.size = size
-        self.gausse_filter()
+        self.gaussian_filter()
         self.increase_contrast()
-        self.crop_to_letter()
-        self.resize()
+        #self.crop_to_letter()
+        #self.resize()
 
     def increase_contrast(self):
         self.data -= self.data.min()
-        self.data /= self.data.max()
+        self.data = self.data.astype(float) / self.data.max()
 
-    def gausse_filter(self):
-        filter = GaussianFilter(1.1)
-        filter.filter(self)
+    def gaussian_filter(self):
+        GaussianFilter(self.blur).filter(self)
 
     def crop_to_letter(self):
         cropper = LetterCropper(0.9)

src/find_svm_params.py

@@ -1,21 +1,54 @@
-C = [2 ** p for p in xrange(-5, 16, 2)]:
-Y = [2 ** p for p in xrange(-15, 4, 2)]
-best_result = 0
+#!/usr/bin/python
+from cPickle import load
+from Classifier import Classifier
+
+C = [float(2 ** p) for p in xrange(-5, 16, 2)]
+Y = [float(2 ** p) for p in xrange(-15, 4, 2)]
 best_classifier = None
 
-learning_set = load(file('learning_set', 'r'))
-test_set = load(file('test_set', 'r'))
+print 'Loading learning set...'
+learning_set = load(file('learning_set.dat', 'r'))
+print 'Learning set:', [c.value for c in learning_set]
+print 'Loading test set...'
+test_set = load(file('test_set.dat', 'r'))
+print 'Test set:', [c.value for c in test_set]
 
 # Perform a grid-search on different combinations of soft margin and gamma
+results = []
+best = (0,)
+i = 0
+
 for c in C:
     for y in Y:
         classifier = Classifier(c=c, gamma=y)
         classifier.train(learning_set)
         result = classifier.test(test_set)
 
-        if result > best_result:
-            best_classifier = classifier
+        if result > best[0]:
+            best = (result, c, y, classifier)
+
+        results.append(result)
+        i += 1
+        print '%d of %d, c = %f, gamma = %f, result = %d%%' \
+              % (i, len(C) * len(Y), c, y, int(round(result * 100)))
+
+i = 0
+
+print '\n     c\y',
+for y in Y:
+    print '| %f' % y,
+
+print
+
+for c in C:
+    print ' %7s' % c,
+
+    for y in Y:
+        print '| %8d' % int(round(results[i] * 100)),
+        i += 1
+
+    print
 
-        print 'c = %f, gamma = %f, result = %d%%' % (c, y, int(result * 100))
+print '\nBest result: %.3f%% for C = %f and gamma = %f' % best[:3]
 
-best_classifier.save('best_classifier')
+best[3].save('classifier.dat')

src/load_characters.py

@@ -0,0 +1,24 @@
+#!/usr/bin/python
+from os import listdir
+from cPickle import dump
+from pylab import imshow, show
+
+from GrayscaleImage import GrayscaleImage
+from NormalizedCharacterImage import NormalizedCharacterImage
+from Character import Character
+
+c = []
+
+for char in sorted(listdir('../images/LearningSet')):
+    for image in sorted(listdir('../images/LearningSet/' + char)):
+        f = '../images/LearningSet/' + char + '/' + image
+        image = GrayscaleImage(f)
+        norm = NormalizedCharacterImage(image, blur=1, size=(48, 36))
+        #imshow(norm.data, cmap='gray')
+        #show()
+        character = Character(char, [], norm)
+        character.get_single_cell_feature_vector()
+        c.append(character)
+        print char
+
+dump(c, open('characters.dat', 'w+'))

src/test_chars.py

@@ -3,7 +3,7 @@ from pylab import subplot, show, imshow, axis
 from cPickle import load
 
 x, y = 25, 25
-chars = load(file('chars', 'r'))[:(x * y)]
+chars = load(file('characters.dat', 'r'))[:(x * y)]
 
 for i in range(x):
     for j in range(y):

src/test_chars_a.py

@@ -1,14 +0,0 @@
-#!/usr/bin/python
-from pylab import subplot, show, imshow, axis
-from cPickle import load
-
-chars = filter(lambda c: c.value == 'A', load(file('chars', 'r')))
-
-for i in range(10):
-    for j in range(3):
-        index = j * 10 + i
-        subplot(10, 3, index + 1)
-        axis('off')
-        imshow(chars[index].image.data, cmap='gray')
-
-show()

src/test_classifier.py

@@ -0,0 +1,68 @@
+#!/usr/bin/python
+from xml_helper_functions import xml_to_LicensePlate
+from Classifier import Classifier
+from cPickle import dump, load
+
+chars = load(file('characters.dat', 'r'))
+learning_set = []
+test_set = []
+
+#s = {}
+#
+#for char in chars:
+#    if char.value not in s:
+#        s[char.value] = [char]
+#    else:
+#        s[char.value].append(char)
+#
+#for value, chars in s.iteritems():
+#    learning_set += chars[::2]
+#    test_set += chars[1::2]
+
+learned = []
+
+for char in chars:
+    if learned.count(char.value) == 70:
+        test_set.append(char)
+    else:
+        learning_set.append(char)
+        learned.append(char.value)
+
+print 'Learning set:', [c.value for c in learning_set]
+print 'Test set:', [c.value for c in test_set]
+print 'Saving learning set...'
+dump(learning_set, file('learning_set.dat', 'w+'))
+print 'Saving test set...'
+dump(test_set, file('test_set.dat', 'w+'))
+#----------------------------------------------------------------
+print 'Loading learning set'
+learning_set = load(file('learning_set.dat', 'r'))
+
+# Train the classifier with the learning set
+classifier = Classifier(c=512, gamma=.125, cell_size=12)
+classifier.train(learning_set)
+classifier.save('classifier.dat')
+print 'Saved classifier'
+#----------------------------------------------------------------
+print 'Loading classifier'
+classifier = Classifier(filename='classifier.dat')
+print 'Loading test set'
+test_set = load(file('test_set.dat', 'r'))
+l = len(test_set)
+matches = 0
+
+for i, char in enumerate(test_set):
+    prediction = classifier.classify(char, char.value)
+
+    if char.value == prediction:
+        print ':-----> Successfully recognized "%s"' % char.value,
+        matches += 1
+    else:
+        print ':( Expected character "%s", got "%s"' \
+                % (char.value, prediction),
+
+    print '  --  %d of %d (%d%% done)' % (i + 1, l, int(100 * (i + 1) / l))
+
+print '\n%d matches (%d%%), %d fails' % (matches, \
+        int(100 * matches / len(test_set)), \
+        len(test_set) - matches)

src/combined_test.py → src/test_combined.py

@@ -1,3 +1,4 @@
+#!/usr/bin/python
 from GrayscaleImage import GrayscaleImage
 from NormalizedCharacterImage import NormalizedCharacterImage
 

src/test_compare.py

@@ -5,32 +5,39 @@ from GrayscaleImage import GrayscaleImage
 from cPickle import load
 from numpy import zeros, resize
 
-chars = load(file('chars', 'r'))[::2]
+chars = load(file('characters.dat', 'r'))[::2]
 left = None
 right = None
 
-for c in chars:
-    if c.value == '8':
-        if left == None:
-            left = c.image
-        elif right == None:
-            right = c.image
-        else:
-            break
+s = {}
 
-size = 16
+for char in chars:
+    if char.value not in s:
+        s[char.value] = [char]
+    else:
+        s[char.value].append(char)
+
+left = s['F'][2].image
+right = s['A'][0].image
+
+size = 12
 
 d = (left.size[0] * 4, left.size[1] * 4)
 #GrayscaleImage.resize(left, d)
 #GrayscaleImage.resize(right, d)
 
 p1 = LocalBinaryPatternizer(left, size)
+h1 = p1.get_single_histogram()
 p1.create_features_vector()
 p1 = p1.features
+
 p2 = LocalBinaryPatternizer(right, size)
+h2 = p2.get_single_histogram()
 p2.create_features_vector()
 p2 = p2.features
 
+total_intersect = h1.intersect(h2)
+
 s = (len(p1), len(p1[0]))
 match = zeros(left.shape)
 m = 0
@@ -52,6 +59,7 @@ for y in range(s[0]):
         m += intersect
 
 print 'Match: %d%%' % int(m / (s[0] * s[1]) * 100)
+print 'Single histogram instersection: %d%%' % int(total_intersect * 100)
 
 subplot(311)
 imshow(left.data, cmap='gray')

src/GaussianFilterTest.py → src/test_gauss.py

@@ -1,3 +1,4 @@
+#!/usr/bin/python
 from GaussianFilter import GaussianFilter
 from GrayscaleImage import GrayscaleImage
 

src/histogram_test.py → src/test_histogram.py

@@ -1,3 +1,4 @@
+#!/usr/bin/python
 from Histogram import Histogram
 
 his = Histogram(10, 10, 110)
@@ -22,4 +23,4 @@ his.add(99)
 his.add(100)
 his.add(109)
 
-print his.bins
+print his.bins

src/LocalBinaryPatternizerTest.py → src/test_lbp.py

@@ -1,3 +1,4 @@
+#!/usr/bin/python
 from GrayscaleImage import GrayscaleImage
 from LocalBinaryPatternizer import LocalBinaryPatternizer
 

src/LetterCropperTest.py → src/test_lettercropper.py

@@ -1,3 +1,4 @@
+#!/usr/bin/python
 from LetterCropper import LetterCropper
 from GrayscaleImage import GrayscaleImage
 
@@ -7,4 +8,4 @@ cropper = LetterCropper(image)
 
 cropped_letter = cropper.get_cropped_letter()
 
-cropped_letter.show()
+cropped_letter.show()