jonahkaye commited on
Commit
6dcd7d1
·
1 Parent(s): bdfbfc5
Files changed (1) hide show
  1. handler.py +2 -1
handler.py CHANGED
@@ -3,6 +3,7 @@ from transformers import LayoutLMForTokenClassification, LayoutLMv2Processor
3
  import torch
4
  from subprocess import run
5
  import pytesseract
 
6
 
7
  # install tesseract-ocr and pytesseract
8
  run("apt install -y tesseract-ocr", shell=True, check=True)
@@ -30,5 +31,5 @@ class EndpointHandler:
30
  # process input
31
  image = data.pop("inputs", data)
32
 
33
- result = pytesseract.image_to_string(image)
34
  return {"predictions": result}
 
3
  import torch
4
  from subprocess import run
5
  import pytesseract
6
+ from pytesseract import Output
7
 
8
  # install tesseract-ocr and pytesseract
9
  run("apt install -y tesseract-ocr", shell=True, check=True)
 
31
  # process input
32
  image = data.pop("inputs", data)
33
 
34
+ result = pytesseract.image_to_string(image, config='--psm 3', output_type=Output.STRING)
35
  return {"predictions": result}