From e0996e35862256114826a6314dc649972972a60c Mon Sep 17 00:00:00 2001
From: Samo Penic <samo.penic@gmail.com>
Date: Sat, 17 Nov 2018 13:42:52 +0000
Subject: [PATCH] Multiple SID robustness..
---
aoiOcr.py | 30 +++++++++++++++++++-----------
1 files changed, 19 insertions(+), 11 deletions(-)
diff --git a/aoiOcr.py b/aoiOcr.py
index 3554c92..9188b0a 100644
--- a/aoiOcr.py
+++ b/aoiOcr.py
@@ -3,11 +3,11 @@
from glob import glob
-settings = {"sid_mask": "64xx0xxx", "answer_treshold": 0.25}
+settings = {"sid_mask": "64xx0xxx", "answer_threshold": 0.25}
classifier = joblib.load("filename.joblib")
#p = Paper(filename="testpage300dpi_scan1.png")
-#p=Paper(filename='sizif111.tif', sid_classifier=classifier, settings={"sid_mask": "11xx0xxx", "answer_treshold": 0.25})
+#p=Paper(filename='sizif111.tif', sid_classifier=classifier, settings={"sid_mask": "11xx0xxx", "answer_threshold": 0.25})
#p=Paper(filename='processed_scans/20141016095134535_0006.tif', sid_classifier=classifier, settings=settings)
#p = Paper(filename="processed_scans/20151111080408825_0001.tif",sid_classifier=classifier,settings=settings,)
#p=Paper(filename='processed_scans/20151028145444607_0028.tif', sid_classifier=classifier, settings=settings)
@@ -20,9 +20,10 @@
"processed_scans/20141021095744144_0009.tif",
"processed_scans/20141028095553745_0018.tif",
"processed_scans/20151013180545275_0011.tif",
- "processed_scans/20160408140801098_0004.tif"
+ "processed_scans/20160408140801098_0004.tif",
+ "processed_scans/20160510075445995_0026.tif"
]
-p=Paper(filename=pa[8], sid_classifier=classifier, settings=settings)
+p=Paper(filename=pa[9], sid_classifier=classifier, settings=settings)
# print(p.QRData)
# print(p.errors)
@@ -35,13 +36,20 @@
print(p.get_paper_ocr_data())
-exit(0)
+
+
filelist = glob("processed_scans/*.tif")
+wrong_sid=0;
+total=0
for f in sorted(filelist):
print("processing: {}".format(f))
- print(
- f,
- Paper(
- filename=f, sid_classifier=classifier, settings=settings
- ).get_paper_ocr_data(),
- )
+ p=Paper(filename=f, sid_classifier=classifier, settings=settings).get_paper_ocr_data()
+ print(f,p)
+ if(p['page_no']==0):
+ total+=1
+ if(len(p['errors'])!=0):
+ wrong_sid+=1
+ if total%10 == 0:
+ print("Total:{}, wrong SID: {}".format(total,wrong_sid))
+
+print("Total:{}, wrong SID: {}".format(total,wrong_sid))
\ No newline at end of file
--
Gitblit v1.8.0