forked from sdaps/sdaps
-
Notifications
You must be signed in to change notification settings - Fork 0
/
README
109 lines (82 loc) · 3.34 KB
/
README
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
SDAPS
=====
This Program can be used to carry out paper based surveys.
The questionnaire is designed using either OpenOffice/LibreOffice and then
exported to PDF. SDAPS then uses both the ODT document and the PDF file
to figure out what questions are asked, and where checkboxes and freeform
fields are placed.
Another great way to use SDAPS is together with its LaTeX class. It allows
you to create questionnaires very easily and is tightly integrated into the
SDAPS main program.
After this, the program can create an arbitrary number of (unique)
questionnaires that can be printed and handed out. After being filled out, you
just scan them in, let sdaps run over them, and let it create a report with
the results.
Requirements
============
Depending on what part of SDAPS you use, different dependencies are
required.
general (including recognize):
* Python 2.7
* distutils and distutils-extra
* python-cairo (including development files)
* libtiff (including development files)
* pkg-config
* python-zbar for "code128" style
* python development files
graphical user interface (gui):
* GTK+ and introspection data
* python-gi
reportlab based reports (report):
* reportlab
* Python Imaging Library (PIL)
ODT based questionnaires (setup/stamp):
* pdftk or pyPdf (pdftk is much faster if you need questionnaire ids)
* python-pdftools
LaTeX based questionnaires (setup_tex/stamp):
* pdflatex and packages:
* PGF/TikZ
* translator (part of beamer)
* and more
LaTeX based reports:
* siunitx
Debug output (annotate):
* Poppler and introspection data
* python-gi
Installation
============
You can install sdaps using "./setup.py install". The C extension will
be compiled automatically, but of course you have to have all the
dependencies installed for this to work.
Standalone execution
====================
As an alternative to installing sdaps it is also supported to run it without
installation. To do this run "./setup.py build" to build the binary modules
and translation. After this execute sdaps using the provided "sdaps.py"
script in the toplevel directory.
Using SDAPS
===========
Please run sdaps with "--help" after installing it for a list of commands.
Also check the website http://sdaps.sipsolutions.net for some examples.
Quality of the recognition
==========================
The quality of the recognition in SDAPS is quite good in my experience.
There is a certain amount of wrong detections, that mostly arise from people
not checking or filling out the boxes correctly. For example:
* The cross is not inside the checkbox, but next to it
* People cross the same box multiple times
* People use very thick pens
* Filling out is not done properly
As you can see, most of the errors arise from the possibility to correct
wrong marks by filling out checkboxes. SDAPS tries to be smart about this
by using different heuristics to detect the case, but it is not foolproof.
Suggestions on how to decrease the error rate are of course welcome.
Matrix Errors
-------------
It can happen that SDAPS is not able to calculate the transformation matrix
which transforms the pixel space of the image into the mm coordinate system
used internally. If this happens the affected pages cannot be further
analysed.
It is usually possible to manually correct them using the GUI, but that can
be quite tedious.
See also TROUBLESHOOTING for some more information.