(Hopefully) fixed data type problems with find_objects/label on 32 bi…

…t machines.

ocrolib/common.py

@@ -50,28 +50,40 @@ def pyargsort(seq,cmp=cmp,key=lambda x:x):
     return sorted(range(len(seq)),key=lambda x:key(seq.__getitem__(x)),cmp=cmp)
 
 def renumber_by_xcenter(seg):
-    objects = [(slice(0,0),slice(0,0))]+measurements.find_objects(seg)
+    objects = [(slice(0,0),slice(0,0))]+find_objects(seg)
     def xc(o): return mean((o[1].start,o[1].stop))
     xs = array([xc(o) for o in objects])
     order = argsort(xs)
     segmap = zeros(amax(seg)+1,'i')
     for i,j in enumerate(order): segmap[j] = i
     return segmap[seg]
 
-def flexible_find_objects(image):
-    """Like measurements.find_objects, but tries to
-    be a bit more flexible about the datatypes it accepts."""
-    # first try the default type
-    try: return measurements.find_objects(image)
+def label(image,**kw):
+    """measurements.label fails to document what types it accepts,
+    and it fails randomly with different types on different
+    platforms.  This tries to work around that."""
+    try: return measurements.label(image,**kw)
     except: pass
-    types = ["int32","int64","int16"]
+    types = ["int32","uint32","int64","unit64","int16","uint16"]
     for t in types:
-        # try with type conversions
-	try: return measurements.find_objects(array(image,dtype=t)) 
+	try: return measurements.label(array(image,dtype=t),**kw) 
 	except: pass
     # let it raise the same exception as before
-    return measurements.find_objects(image)
+    return measurements.label(image,**kw)
 
+def find_objects(image,**kw):
+    """measurements.find_objects fails to document what types it accepts,
+    and it fails randomly with different types on different
+    platforms.  This tries to work around that."""
+    try: return measurements.find_objects(image,**kw)
+    except: pass
+    types = ["int32","uint32","int64","unit64","int16","uint16"]
+    for t in types:
+	try: return measurements.find_objects(array(image,dtype=t),**kw) 
+	except: pass
+    # let it raise the same exception as before
+    return measurements.find_objects(image,**kw)
+
 def rgb2int(image):
     """Converts a rank 3 array with RGB values stored in the
     last axis into a rank 2 array containing 32 bit RGB values."""
@@ -109,7 +121,7 @@ def setImageMasked(self,image,mask=None,lo=None,hi=None):
         labels,correspondence = renumber_labels_ordered(labels,correspondence=1)
         self.labels = labels
         self.correspondence = correspondence
-        self.objects = [None]+flexible_find_objects(labels)
+        self.objects = [None]+find_objects(labels)
     def setPageColumns(self,image):
         """Set the image to be iterated over.  This should be an RGB image,
         ndim==3, dtype=='B'.  This iterates over the columns."""
@@ -811,7 +823,7 @@ def estimate_xheight(line,scale=1.0,debug=0):
     return bottom-top,bottom
 
 def keep_marked(image,markers):
-    labels,_ = measurements.label(image)
+    labels,_ = label(image)
     imshow(sin(17.1*labels),cmap=cm.jet)
     marked = unique(labels*(markers!=0))
     print marked
@@ -840,7 +852,7 @@ def latin_filter(line,scale=1.0,r=1.2,debug=0):
 
 def remove_noise(line,minsize=8):
     bin = (line>0.5*amax(line))
-    labels,n = measurements.label(bin)
+    labels,n = label(bin)
     sums = measurements.sum(bin,labels,range(n+1))
     sums = sums[labels]
     good = minimum(bin,1-(sums>0)*(sums<minsize))

ocrolib/docproc.py

@@ -6,8 +6,8 @@
 from scipy import stats
 from scipy.ndimage import measurements
 from pylab import *
-
 from common import *
+import common
 
 def avg(*args):
     return mean(args)
@@ -19,7 +19,7 @@ def seg_boxes(seg,math=0):
     coordinates are used (however, the order of the values in the
     tuple doesn't change)."""
     seg = array(seg,'uint32')
-    slices = measurements.find_objects(seg)
+    slices = common.find_objects(seg)
     h = seg.shape[0]
     result = []
     for i in range(len(slices)):
@@ -117,7 +117,7 @@ def bbox(image):
     """Compute the bounding box for the pixels in the image."""
     assert len(image.shape)==2,"wrong shape: "+str(image.shape)
     image = array(image!=0,'uint32')
-    cs = scipy.ndimage.measurements.find_objects(image)
+    cs = common.find_objects(image)
     if len(cs)<1: return None
     c = cs[0]
     return (c[0].start,c[1].start,c[0].stop,c[1].stop)

ocrolib/grouper.py

@@ -69,7 +69,7 @@ def setSegmentation(self,segmentation,cseg=0,preferred=None):
             # print sorted(correspondences)
             self.pre2seg = correspondences
         # compute the bounding boxes in order
-        boxes = [None]+measurements.find_objects(segmentation)
+        boxes = [None]+common.find_objects(segmentation)
         n = len(boxes)
         # now consider groups of boxes
         groups = []
@@ -107,7 +107,7 @@ def setCSegmentation(self,segmentation):
         the groups corresponding to each labeled object.  Objects should be labeled
         consecutively."""
         # compute the bounding boxes in order
-        boxes = [None] + measurements.find_objects(segmentation)
+        boxes = [None] + common.find_objects(segmentation)
         n = len(boxes)
         # now consider groups of boxes
         groups = []

ocrolib/lineseg.py

@@ -1,6 +1,7 @@
 from pylab import *
 from scipy.ndimage import filters,morphology,measurements
 import psegutils
+import common
 
 
 
@@ -85,7 +86,7 @@ def ccslineseg(image):
     center = filters.maximum_filter(center,(3,3))
     center = psegutils.keep_marked(image>0.5,center)
     center = filters.maximum_filter(center,(2,2))
-    center,_ = measurements.label(center)
+    center,_ = common.label(center)
     center = psegutils.spread_labels(center)
     center *= image
     return center
@@ -136,7 +137,7 @@ def charseg(self,line):
         tracks = dplineseg2(line,imweight=self.imweight,bweight=self.bweight,
                             diagweight=self.diagweight,debug=self.debug,r=self.r)
         tracks = array(tracks<0.5*amax(tracks),'i')
-        tracks,_ = measurements.label(tracks)
+        tracks,_ = common.label(tracks)
         self.tracks = tracks
         rsegs = psegutils.spread_labels(tracks)
         rsegs = rsegs*(line>0.5*amax(line))

ocrolib/psegutils.py

@@ -4,6 +4,7 @@
 from scipy.ndimage import filters,interpolation,morphology,measurements
 from scipy import stats
 from scipy.misc import imsave
+import common
 
 class record:
     def __init__(self,**kw): self.__dict__.update(kw)
@@ -80,7 +81,7 @@ def spread_labels(labels,maxdist=9999999):
     return spread
 
 def keep_marked(image,markers):
-    labels,_ = measurements.label(image)
+    labels,_ = common.label(image)
     marked = unique(labels*(markers!=0))
     kept = in1d(labels.ravel(),marked)
     return (image!=0)*kept.reshape(*labels.shape)
@@ -100,7 +101,7 @@ def correspondences(labels1,labels2):
 
 def propagate_labels_simple(regions,labels):
     """Spread the labels to the corresponding regions."""
-    rlabels,_ = measurements.label(regions)
+    rlabels,_ = common.label(regions)
     cors = correspondences(rlabels,labels)
     outputs = zeros(amax(rlabels)+1,'i')
     for o,i in cors.T: outputs[o] = i
@@ -109,7 +110,7 @@ def propagate_labels_simple(regions,labels):
 
 def propagate_labels(regions,labels,conflict=0):
     """Spread the labels to the corresponding regions."""
-    rlabels,_ = measurements.label(regions)
+    rlabels,_ = common.label(regions)
     cors = correspondences(rlabels,labels)
     outputs = zeros(amax(rlabels)+1,'i')
     oops = -(1<<30)
@@ -126,8 +127,8 @@ def A(s): return W(s)*H(s)
 def M(s): return mean([s[0].start,s[0].stop]),mean([s[1].start,s[1].stop])
 
 def binary_objects(binary):
-    labels,n = measurements.label(binary)
-    objects = measurements.find_objects(labels)
+    labels,n = common.label(binary)
+    objects = common.find_objects(labels)
     return objects
 
 def estimate_scale(binary):
@@ -153,7 +154,7 @@ def compute_boxmap(binary,scale,threshold=(.5,4),dtype='i'):
 def compute_lines(segmentation,scale):
     """Given a line segmentation map, computes a list
     of tuples consisting of 2D slices and masked images."""
-    lobjects = measurements.find_objects(segmentation)
+    lobjects = common.find_objects(segmentation)
     lines = []
     for i,o in enumerate(lobjects):
         if o is None: continue
@@ -291,8 +292,8 @@ def rgbshow(r,g,b=None,gn=1,cn=0,ab=0,**kw):
     imshow(clip(combo,0,1),**kw)
 
 def select_regions(binary,f,min=0,nbest=100000):
-    labels,n = measurements.label(binary)
-    objects = measurements.find_objects(labels)
+    labels,n = common.label(binary)
+    objects = common.find_objects(labels)
     scores = [f(o) for o in objects]
     best = argsort(scores)
     keep = zeros(len(objects)+1,'B')