What is the News?
Here we
present the differences, improvements and
new tools that were
implemented in the 5.5.3 SPRING Version.
What's New in Spring 5.5.3
New
classifiers have been added to in the
version 5.5.3, unsupervised
based on Maximization and Expectations and
the Classifier by Decision tree. Decision
trees are statistical models that use
supervised training for classification. In
other words, in its construction is used a
training set formed by inputs and outputs of
the classes themselves.
What's New in
Spring 5.5.2
- REGISTRATION
OF PUNCTUAL IMAGES, BY
LINES, AUTO AND SEMI
AUTOMATIC WITH VALIDATIONS
In
version
5.5.2
of SPRING,
the new tool for Automatic
and Semiautomatic Image Registration
has been added. With the use of the
Terralib5.2 library, it is possible to use
points of interest localization by Moravec
or by SURF, also using new
Geometric Transformations that the new
library has. Such as "Afinne", "Projective", "RST - (Resampling, Scaling and
translation)" among others.
The new registry also makes use of Multi-Thread to optimize the
location of control points.
Another utility
added in the tool was the validation of
image registration information with a
positional accuracy measurel of points and
features, whose function is to evaluate
positioning accuracies through specific
sampling points or features. Precision
evaluation is done using known
trigonometric functions or statistics (eg
RMS (Mean Mean Square deviation), mean and
standard deviation among others, with
absolute or relative values.
What's New in
Spring 5.5.1
In
SPRING
version 5.5.1, the new Image
Fusion toolhas been
added. The Image
Fusion tecnique
combines the colors of the low-resolution
multispectral images with the details of a
panchromatic band that has better spatial
resolution. The methods added were "IHS
Fusion", based on the IHS
transform, ,
" PCA Fusion", based on the Principal
Component Analysis
transform and
the "Wisper Fusion" based on the Wavelet
transform.
What's New in
Spring 5.5.0
- AUPDATING, UNIFYING AND
COMPATIBILITY OF THE TERRALIB
LIBRARY 5.2
In version
5.5.0 do SPRING, improvements were added
for users, with emphasis on updating,
unification and compatibility with the Terralib 5.2 library.
In this way is possible
to use functions from Terralib 5.2 library already update
in the SPRING environment with the
guarantee of greater integration of the
data stored between the two platforms.
The new database
converter is a tool intended to make it
easier for the user to convert data
structures from the SPRING format to Terralib 5.2 default.
One of the most important
features is that SPRING 5.5.0 is the
integration capability with
object-relational database management
systems (RDBMS-OR) in the spatial data
store. This integration will allow, for
example, the sharing of large databases in
corporate environments by custom
applications for different types of users.
In SPRING 5.5.0 new tools
were added in the process of segmentation
and extraction of information from remote
sensing images regardless of their
characteristics (high or low spatial
resolution, for example).
The
improvement of the spatial and spectral
resolution of the sensors in the last
years evidenced the difficulty of applying
the classical segmenters, which extracted
information at the pixel level, mainly in
remote sensing data of low spatial
resolution. In these images a single pixel
can store the spectral response of more
than one object, or a heterogeneous area
of the same object. In high resolution
images, a pixel has similar spectral
characteristics to its neighbors, being
purer in terms of spectral mixing.
Due to the
difficulty of extracting "homogeneous"
regions mainly from high resolution
images, the version 5.5.0 of SPRING incorporates
the Baatz
Segmenter which is
oriented to multiscale
and multiresolution. The procedure for
multiple-scale image segmentation can be
described as a region fusion technique. It
starts with each pixel forming an image
object (region). At each iteration, each
object merges with a neighboring object.
This neighbor is the one for which the
object resulting from the fusion
represents the smallest increase in
heterogeneity in relation to the sum of
the internal heterogeneity measures of the
two fusion candidate objects. The merger
decision is based on criteria of local
homogeneity. The difference between the
heterogeneities calculated for the two
moments is that it defines the short of
the fusion of the two objects.
For more
informations about this method access: http://www.ecognition.cc/download/baatz_schaepe.pdf
This new
approach intends to approximate the
techniques of processing and extracting
information with human reasoning, that is,
with the way of thinking and recognizing
patterns that we intuitively adopt when
interpreting an image or appreciating a
landscape (BAATZ e SCHÃPE, 2000).
Another
segmenter was built into this release.
This is the Multi Region Segmenter. It
follows the same principles as SPRING´s native
Region Segmenter, but exploits
multiprogramming, multithread, and
multicore features. For further
explanations of these characteristics
access:
Multicore: https://pt.wikipedia.org/wiki/Multin%C3%BAcleo
Multiprogramação:https://www.ime.usp.br/~gold/cursos/2002/mac2301/ep2/ep2/node3.html
Multithread
https://pt.wikipedia.org/wiki/Thread_(ci%C3%AAncia_da_computa%C3%A7%C3%A3o)
Segmenters
from previous versions have been retained
and can be used mainly in low and medium
spatial resolution remote sensing data.
What's New in
Spring 5.4.x
In
the version 5.4 of SPRING,
the tool has been modified to allow the
processing of image restoration of the
satellite sensors CBERS-4
and
LANDSAT-8
apart of the sensors already included in
the previous versions.
In the SPRING 5.4 was
included the options to import and export
images from the BigTIff , format,
which allows it to store images in larger
physical files and with more color
definition, since the conventional TIFF is
limited to 4GB.
In the BigTiff format,
it’s possible to store images larger than
4GB in 64 bits. Actually a Landsat-8
image, for example, with all bands added
can exceed 1,65 GB knowing that it is a
medium spatial resolution. A high
resolution image easily exceed 4 GB.
For more
informations about BigTiff access:
http://bigtiff.org/
- Update
of external libraries
In
the SPRING 5.4 the
external libraries were upgraded to general
improveme of performance and stability with
many available geoprocessing and sensoring
platforms.
The
updated libraries were:
- Libtiff;
- Libqwt;
- Libgeotiff;
- Libshape;
- LibDXF;
- Libgeosc;
- Libfreetypes;
- Libjpeg;
- Libjpeg2000.
The
Terralib libraries were also updated for
the most recent version of the OGC (Open
Geoespatial Consortium) specifications.
The
existing librariers for the version 5.4 of
SPRING in Linux, were updated to run the Ubuntu 16.04, version,
all the specifications of the OGC (Open
Geoespatial Consortium) and its
technologies were update as well.
From
version 5.5.0, there was also a
replacement of the Toolkit Qt for the
version 5.8.
This
tool has been native to SPRING since the
early versions. With the tool it is possible
to delimit or create a mask in any infolayer
of a project and to clip it.
One of the
news of the SPRING 5.4 is the
possibility to choose many areas of a
polygon as mask, and with that you can
simultaneous clip a information plane,
since previously the tool allowed the clip
of a single polygon at once.
It is the
optimization of the operation of infolayer
clipping, since it becomes possible to
execute multiple cuts in a single
infolayer.
In
order to make SPRING more
compatible with the availables data formats,
it was added the Import and Export operations
in CSV (Comma
Separed Values).
The CSV format
allows to store statistics, social and
environmental data in lines and columns,
separated by a delimiter. The CSV files
made are connected to a map via SPRING and
from this connection, the polygons, lines
or points are associated to a label, that
associates the presented attributes in
this table.
The CSV format is
supported by various available
spreadsheets and SGDBs.
In the
version SPRING 5.4 was
incorporated the SQLSERVER,
relational database manager system, among
its main features are:
- It is available
for 32 and 64 bit environments;
- It is available
for Windows or Linux operational
systems;
- It does not
require the installation of a database
manager driver for its use;
- It provides a
viewer to external access to the data
tables.
What's New in Spring 5.2
- Orthorectification
In
order to improve the geometric correction of the
images used in the SPRING
5.2, it
was added a tool for image orthorectification.
The orthorectification,
also called 3D geometric correction, aims to
reduce distortions,
introduced by the attitude of the sensor during
the image acquisition,
as well as errors due to relief displacement.
The orthorectification
tool available in
the SPRING
5.2 is
based on the Model of Rational Functions and
uses information contained
in RPC (Rational Polynomial Coefficients) files
and a Digital Elevation
Model, to remove systematic errors caused by the
platform, terrain and
sensor.
The
orthorectification procedure allows to obtain an
image in orthogonal
perspective that has better accuracy and
provides better results when
it is used as a data source to derive
cartographic documents.
- Vectorization
It
was developed, for SPRING
5.2 a vectorization
tool that enables a
semi-automatic, or automatic, extraction of a
vector data from an input
image.
This
tool can assist the user in the process of
vector editing, optimizing
working time. The user can, for example,
automatically obtain the
vector information from a raster topographic map
from the Brazilian
Institute of Geography and Statistics (IBGE).
Two
options are available for vectorization:
automatic and
semi-automatic. In automatic mode the
vectorization process occurs
without user intervention. In the
semi-automatic, available only inside
the Vector Edit tool, you can perform a partial
image vectorization.
- OGC
With SPRING version5.2 internet online maps
can be displayed in its main visualization
window.
For
this, it was deployed two modules in the system
that allows to access,
or to provide, remote data available from WMS
(OpenGIS Web Map
Service) and WFS
(Open Geospatial
Consortium Web Feature Service Interface
Standard),making the
SPRING 5.2
client toWMS and
WFS servers.
- SCARTA
The SCARTA,module
was
restructured for this SPRING new
version.
Now the user can access charts, or its frames,
directly from
the system interface in order to display or to
edit them.
The
new SCARTA
allows access to its
all functions in a practical way. It allows also
using the various
tools built into
SPRING in
its updating processes as, for example, the new
control panel and
automatic display of information layers.
- TerraLib
In order to provide
greater
integration of data fromTerraLib database with the SPRING
5.2 a
visualization module forTerraLib databases
(themes
and views) was implemented in this SPRING
5.2.This
tool also provides an option for saving the Terralib information in aSPRING
5.2
database.
- Plugins
The plug-in
echnology allows
external software developers to create new tools
for a software system
providing to it new specific functionalities.
In
the SPRING
5.2 was
developed a function that allows the user to add plugins.
Therefore the user
now will be able to add new functionalities to the
SPRING according to
his needs. Any user also will be able to turn
available in the system
new functionalities that he seems it is important
to share with other
SPRING users.
As
an example for this new version, we have a plug-in
hat makes it
possible to access or provide remote data
available from WMS and WFS
services.
- SQLITE
The
database manager SQLITE
was incorporated
into SPRING
with the aim of
replacing the database manager Dbase,
because it had such
limited support for 64 bit and it is not
freeware.
The SQLITE
operation is similar
to the Dbase. Among its main features are:
- It is available
for 32 and 64 bit
environments;
- It is available
for Windows or Linux
operational systems;
- It does not
require the installation
of a database manager driver for its use;
- It has its source
code available for
free;
- It provides a
viewer to external
access to the data tables.
Others
Thinning
- Implementation of thinning tool for raster data.
Raster
maps resulting from a scanner process, over a map
sheet, frequently
present lines with widths larger than one pixel.
This, of course,
depends on the resolution of the scanner being
used. In automatic
vectorization process from raster data, this
thick, (“fat”), lines are
transformed in misleading polygons. This leads to
a further forced
vector edition task to have better line
representations. To avoid this
type of errors, the thinning process thin the
thick lines trying to
maintain their widths equal 1 pixel. The thinning
process can be used
as a preprocessing to create more reliable vector
representations of
scanned raster maps.
Assistant
Database -
Wizard for database creation add in the plug-in
technology.
This
tool was developed to assist the user in creating
a database on SPRING
5.2, making this task easier and more intuitive.
Restoration
of CBERS 2B
The
image restoration is a radiometric correction
technique used to correct
distortions inserted by the optical sensor during
the process of
digital images acquisition. In order to keep the
SPRING 5.2 updated,
the Restore tool was modified to process images of
the available
sensors on the satellite CBERS 2B.
Resampling
A
tool for resampling image was added in order to
generate, in a simple
way, an output image with a different pixel size
value, defined by the
user, of the input image.
Line
Simplification
A
line simplification tool was added to the models
of SPRING vector 5.2.
The process of line simplification, previously
available only through
mosaic tool, is now available in the menu of
vector data models.
|