We use the sqlite database system which basically stores the database in a single file (and doesn't follow the client-server model).
The data is stored in valsketch/testing.db
which must be created by the user. Then, to create the relevant tables within the
file:
python manager.py createdb
This would ensure that the database is now ready for use to store the sketch annotations.
We recommend the use of virtual environments. Once a virtual environment is setup and activated, you should install all packages in listed
in the requirements.txt
file by pip install -r requirements.txt
. This command can be executed even outside a virtual environment, but be aware
that it will install these python packages globally and may require admin privileges.
Since the tool was made to crowdsource sketch annotations, each user of the tool is limited to a fixed number of sketches to annotate. This number
can be specified in the valsketch/static/idx.txt
file.
This annotator was made to annotate parts of sketches of different objects. For each object, a number of parts are listed in valsketch/static/parts.json
file. You may edit the file to suit your needs. Note that the object name that is used in the parts.json
file must match the directory name which stores
sketches of that particular object.
All sketches must be stored in suitably named directories in valsketch/static/sketches_svg
directory. The annotator tool only works with .svg
files so make
sure that sketches are of this format.