ocropus-ldewarp

#!/usr/bin/python

# make the clustering class-specific

import sys,os,re,glob,math,glob,signal,traceback,codecs,time
import fcntl
import matplotlib,tables
if "DISPLAY" not in os.environ: matplotlib.use("AGG")
else: matplotlib.use("GTK")
from scipy.ndimage import interpolation
from pylab import *
from optparse import OptionParser
from multiprocessing import Pool
import ocrolib
from ocrolib import number_of_processors,die
from ocrolib.ligatures import lig
from ocrolib.lineest import TrainedLineGeometry,make_normalizer
from ocrolib import lineest
from ocrolib import lineseg,morph,linerec,improc,lineproc,h5utils
from ocrolib.toplevel import *

import warnings,numpy
warnings.simplefilter('ignore',numpy.RankWarning)
warnings.filterwarnings('error','.*invalid value.*')

signal.signal(signal.SIGINT,lambda *args:sys.exit(1))

import argparse
parser = argparse.ArgumentParser(description = """Dewarps textlines and rescales them to a standard x-height.""")
parser.add_argument("-X","--exec",dest="execute",help="execute before anything else (usually used for imports)",default="None")
parser.add_argument('-d',"--baselinedegree",type=int,default=1,help="degree of baseline model")
parser.add_argument('-e','--lineest',default=ocrolib.default.lineest,help="line geometry model (%(default)s)")
parser.add_argument('-l','--height',default=32,type=int,help="target line height")
parser.add_argument('-x','--xheightfrac',default=0.5,type=float,help="target xheight frac")
parser.add_argument('-b','--baselinefrac',default=0.8,type=float,help="target baseline frac")
parser.add_argument('files',nargs="*",default=[],help="text line files")

args = parser.parse_args()
args.files = ocrolib.glob_all(args.files)
if len(args.files)==0:
    parser.print_help()
    sys.exit(0)

target_height = args.height
exec args.execute
print "# target height",target_height

lnorm = lineest.load_normalizer(args.lineest)
print "lnorm",lnorm

def process1(fname):
    base,_ = ocrolib.allsplitext(fname)
    print "===",fname
    limage = ocrolib.read_image_gray(fname)
    lnorm.measure(amax(limage)-limage)
    limage = lnorm.normalize(limage,cval=amax(limage))
    ocrolib.write_image_gray(base+".dew.png",limage)
    if os.path.exists(base+".cseg.png"):
        cseg = ocrolib.read_line_segmentation(base+".cseg.png")
        cseg = lnorm.normalize(cseg,order=0,dtype=dtype('i'),cval=0)
        ocrolib.write_line_segmentation(base+".dcs.png",cseg)



for fname in args.files:
    try:
        process1(fname)
    except ocrolib.CheckWarning as warning:
        print warning
    except:
        traceback.print_exc()