Convert Georeferenced imagery to vector data using python

Question

Are there any opensource tools I could use to take a projected georeferenced image such as a jpg with world file, that is UTM projection such as NAD83 and convert it into distinct vector polygons using python or possibly just some software? The end result would likely need to be a shp file.

Answer 1

The Gdal_polygonize.py that is shipped with GDAL library should do the job.

Answer 2

It appears you can do the conversion in QGIS using GRASS tools:

Once we have a new QGM configuration file, we can add the module to the list of modules. The list is defined in qgis/src/plugins/grass/config/default.qgc file. The sections in the file are organized hierarchically and represented in menu as tree.

Here is an example of a record for r.to.vect.line.qgm module

section label="Convert raster to vector" grass name="r.to.vect.line.qgm"

For section/module it is possible to specify GRASS version for which it is

valid using version_min/version_max attributes:

version_min - minimum GRASS version for which this section/module is valid version_max - maximum GRASS version for which this section/module is valid General rules for new modules GRASS Tools are intended mostly for beginners and not very experienced users. The basic rules are:

Each module should have 3 or less options; if it seems to be necessary to add more options, it should be discussed beforehand Module definition should be robust so that it doesn't allow a user to run a module with wrong option. If there is not yet support for certain type of option the module should not be added to menu

http://hub.qgis.org/projects/quantum-gis/wiki/Adding_New_Tools_to_the_GRASS_Toolbox