sporada/ocr
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/ocr.git synced 2025-01-30 21:59:01 +00:00
Code Issues Projects Releases Wiki Activity

Start one ocropus-nlbin job per page instead of one per document

Browse Source
This commit is contained in:
Patrick Jentsch 2019-04-11 11:50:09 +02:00
parent 2648f40b92
commit a947e36997
1 changed files with 9 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
ocr
View File

@ -116,15 +116,18 @@ class OCRWorkflow(WorkflowRunner):
# Task "ocropus_nlbin_job": binarize tiff files from previous split
# Dependencies: split_jobs
###
self.waitForTasks()
ocropusnlbin_jobs = []
ocropusnlbin_job_number = 0
for job in self.jobs:
ocropusnlbin_job_number += 1
cmd = 'ocropus-nlbin -o "%s" "%s"/*' % (
os.path.join(job["output_dir"], "tmp", "ocropus-nlbin"),
os.path.join(job["output_dir"], "tmp", "tiff_files")
)
ocropusnlbin_jobs.append(self.addTask(label="ocropusnlbin_job_-_%i" % (ocropusnlbin_job_number), command=cmd, dependencies=split_jobs))
# This list is empty if you don't wait for ocropus_nlbin_jobs to complete
for file in os.listdir(os.path.join(job["output_dir"], "tmp", "tiff_files")):
ocropusnlbin_job_number += 1
cmd = 'ocropus-nlbin -o "%s" "%s"/*' % (
os.path.join(job["output_dir"], "tmp", "ocropus-nlbin"),
os.path.join(job["output_dir"], "tmp", "tiff_files", file)
)
ocropusnlbin_jobs.append(self.addTask(label="ocropusnlbin_job_-_%i" % (ocropusnlbin_job_number), command=cmd, dependencies=split_jobs))
###
# Task "tesseract_job": perform OCR on binarized images