Initial commit.

.hgignore

@@ -0,0 +1,6 @@
+syntax: glob
+
+*~
+*.pyc
+*.swp
+.ipynb_checkpoints

pixelocr/image.py

@@ -0,0 +1,78 @@
+from io import BytesIO
+
+import numpy as np
+from skimage.io import imread, imsave
+from skimage.color import rgb2gray
+
+from .utils import cached_property
+
+
+class Image(object):
+    def __init__(self, data):
+        self._data = data
+
+    def __getitem__(self, key):
+        return type(self)(self._data.__getitem__(key))
+
+    @classmethod
+    def fromfile(cls, filename):
+        return cls(imread(filename))
+
+    @classmethod
+    def space(cls, height, width):
+        return cls(np.ones((height, width)))
+
+    @property
+    def shape(self):
+        return self._data.shape
+
+    @property
+    def T(self):
+        return type(self)(self._data.swapaxes(0, 1))
+
+    def _repr_png_(self):
+        buf = BytesIO()
+        imsave(buf, self._data)
+        return buf.getvalue()
+
+    @property
+    def height(self):
+        return self._data.shape[0]
+
+    @property
+    def width(self):
+        return self._data.shape[1]
+
+    @cached_property
+    def bitmap(self):
+        grayscale = rgb2gray(self._data)
+        return (grayscale < 1).astype('b')
+
+    @property
+    def key(self):
+        height, width, *_ = self.shape
+        shape = '{}x{}'.format(height, width)
+        bitmap = np.packbits(self.bitmap).tostring()
+        return shape.encode('latin1') + b':' + bitmap
+
+    def unframe(self, width=2):
+        return self[width:-width,width:-width]
+
+    def iter_lines(self, min_space=200):
+        line_start = None
+        prev_line_end = 0
+        for i, row in enumerate(self.bitmap):
+            if row.any():  # non-blank row
+                if line_start is None:
+                    line_start = i
+                    if line_start - prev_line_end >= min_space:
+                        yield Image.space(line_start - prev_line_end, self.width)
+            else:
+                if line_start is not None:
+                    yield self[line_start:i, :]
+                    line_start = None  
+                    prev_line_end = i
+
+    def iter_letters(self, min_space=10):
+        for letter in self.T.iter_lines(min_space=7):
+            yield letter.T

pixelocr/utils.py

@@ -0,0 +1,15 @@
+import functools
+
+def cached_property(fun):
+    """A memoize decorator for class properties."""
+    @functools.wraps(fun)
+    def get(self):
+        try:
+            return self._cache[fun]
+        except AttributeError:
+            self._cache = {}
+        except KeyError:
+            pass
+        ret = self._cache[fun] = fun(self)
+        return ret
+    return property(get)