Hi Nishad,

If you are going to classify Landsat Data, you should do a Supervised
Classification. This means that you provide input points, whose Landuse you
know.  This is an iterative method, where you need to provide inputs for
the algorithm to give a better output.

The Link that you shared does Un-Supervised Classification, i.e. without
any apriori knowledge, and hence the output quality will be lower.

I'll suggest you have a look at the Semi-Automatic Classification Plugin
<https://fromgistors.blogspot.com/p/semi-automatic-classification-plugin.html>
for QGIS. The website has excellent video tutorials which show you how you
can achieve a decent LULC classification.

Regards,
Devdatta

On Tue, Sep 12, 2017 at 9:18 AM, nishadh <[email protected]> wrote:

> Hi,
>
> Land use land cover classification for India for the year 1985-2005 based
> on Landsat satellite imagery is available in public domain as geotiff
> imagery https://daac.ornl.gov/VEGETATION/guides/Decadal_LULC_India.html
>
> To get the latest year land use land cover statistics such as of 2015,
> requires need of classification on raw images. Python with its open source
> libraries such as rasterio, numpy and scipy is apt to carry out
> classification. There is a tutorial describing the various steps to address
> using Python https://www.machinalis.com/blog/python-for-geospatial-
> data-processing/
>
> However assessing the classification accuracy and result comparable with
> above open data is no where near to expectations. Requesting guidance in
> this regard.
>
> Regards,
>
> Nishadh.K.A.
>
> --
> Datameet is a community of Data Science enthusiasts in India. Know more
> about us by visiting http://datameet.org
> ---
> You received this message because you are subscribed to the Google Groups
> "datameet" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
Datameet is a community of Data Science enthusiasts in India. Know more about 
us by visiting http://datameet.org
--- 
You received this message because you are subscribed to the Google Groups 
"datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to