Road to the Semi-Automatic Classification Plugin 6 (Greenbelt): the SCP dock

As I wrote this post, QGIS 3 is under development and it should be released in December 2017.
This important QGIS update brings many new features; in particular, the upgrade to Python 3 and QT 5 imply that all the plugins must be updated to these changes.

I am updating the Semi-Automatic Classification Plugin (SCP) to version 6 (codename Greenbelt) which will work with QGIS 3 only (not QGIS 2 because of the API changes).
The codename Greenbelt (Maryland, USA) is the location of the NASA’s Goddard Space Flight Center that had a key role in Landsat satellite development and will lead the Landsat 9 development of the space and flight segments (Landsat 9 to be launched in 2020).

I have already released an SCP experimental version 5.99 that runs in QGIS 3. However, the QGIS 3 development is still going on and other API breaks could cause issues in SCP. Therefore, in order to focus on plugin development, I won't update the plugin repository until the feature freeze of QGIS 3.
The final SCP version 6 will probably be ready by December 2017, in line with the QGIS 3 release.

I am willing to improve the functionalities of SCP with this new version 6, and I'll try to periodically post the development progress.
First, I have updated the SCP dock in order to be more productive and user friendly.

SCP dock

The tab Home includes the buttons for opening the SCP functions, with a section News where updates and news about SCP and related services will be shown on startup.
Other tabs include the Training input commands and the Classification commands.
In addition, several buttons are always visibile in the left part of the dock, easing the launch of other SCP tools.

Training input

The training input allows for the creation of the training input file (.scp ) which is required for storing ROIs and spectral signatures for image classification. The loaded training input is always visible in the top of the dock.
I've tried to maximize the space for the ROI Signature list.
A new option allows you to automatically save the ROI Signature list to the training input file every time a ROI is saved (with the current version of SCP, the ROI Signature list is saved only when the QGIS project is saved).

Macroclass list

An entire tab is dedicated to the macroclass list which is saved into the training input file.

Classification

In the same fashion, classification commands are placed in a new tab.
I have changed the threshold behavior: if the threshold value is greater than 0 , this value is used for all the spectral signatures; otherwise, thresholds for each signatures can be defined in the tab Signature threshold.

This new SCP dock should be more usable also with small screen devices.
I am continuing the development of the SCP 6 (in particular the Band set tool), and in the next few weeks I am going to post other news.

For any comment or question, join the Facebook group and the Google+ Community about the Semi-Automatic Classification Plugin.

Newer posts Older posts