| Some web search engines index only the first
part of long documents. In order to have this document indexed
fully, it was divided into shorter pieces. |
1
2
3
4
5
6
7
8
9
|
| If you have
arrived here from a web search, go to the full document ... |
View in PDF
format
2D
Display.
Direct Display of ESRI Shapefiles.
Linked Via Shape Objects.
Shapefiles are now widely used as a
public means to transfer graphical geodata and associated attributes especially
on the Internet. As a result, many
software products of varying quality are creating them.
More and more shapefiles are being encountered that do not conform to the
published ESRI specifications but may work in an ESRI product due to
undocumented flexibility, features, or adjustments.
Changes to the TNT products
to detect and accommodate these non-standard features in shapefiles are made as
they are encountered, and these adjustments become available to you through the
weekly patches to the TNT products.
However, shapefiles are now being encountered that can not be used in the
appropriate ESRI product. Solutions
to these cases are not guaranteed and they may need to be returned to their
source for improvement.
RV6.9
permitted the selection of shapefiles as layers for use in displays or importing
including their area, line, and point styling.
As noted in the section above entitled Shape Objects, the direct
display capability for shapefiles is now handled via a link through a TNT
shape object rather than a CAD object. This
and associated changes have enabled several new and useful features to be
supported for linked shapefiles. These
features are all enabled via the link file and the original shapefile is not
altered and can be used with its original characteristics and limitations in an
ESRI or other product. Shapefiles
can still be imported and converted to vector objects with topology or CAD
objects. As yet they can not be
imported to an internal shape object since the use of this new object type in
other processes is not yet fully supported.
The shape object Project File created for
this link is merely a shell. It
contains the equivalent TNT styles,
font identities or substitutions, additional tables, table indexes, and other
reference information to make this external shapefile or an Oracle Spatial Layer
appear to all TNT processes as if it
was an internal shape object in a Project File.
(For a more detailed discussion of the use of the shape object as a link
to Oracle Spatial layers see the MicroImages MEMO entitled Release of RV6.9
of the TNT Products.) This is
the same approach used in the transparent linking to and direct display of
images in JPEG, JPEG2000, MrSID, and other raster formats.
Just as for shape objects, these links make the external raster appear to
the TNT processes as if it were
actually an internal raster object.
Faster Links.
The
most significant benefits of linking to shapefiles via a shape object is that it
is much faster as it does not import or change any of the contents of the
shapefile. The previous linking in V6.9
via a CAD object was slow as it had to reconcile element and attribute
associations. In RV7.0
selecting a shapefile as a layer for a complex view will now create its link
sufficiently fast so as to make it unnoticeable that a link is actually being
built. And, with the addition of
the new legend for the shape elements and the standard TNT
features such as map projection conversion, the direct use of shapefiles is
quite transparent! Shapefiles can
still be converted to vector or CAD objects by importing them.
Coordinate Reference System.
Shapefiles
may, or may not be accompanied by the ESRI projection file (*.prj).
This is the problem with having many component files for your project.
If this file is not present, then the shapefile may be in latitude and
longitude coordinates or may not have any projection.
When a *.prj file is available, it is used to establish the Coordinate
Reference System (CRS) in the shape link file.
If no *.prj is present, then it is assumed that the shapefile is in
latitude and longitude coordinates as long as its coordinates are in the range
of latitude and longitude (which means, +/-90 and +/-360, respectively).
If no *.prj is provided and the coordinates exceed these limits, then the
shapefile is assumed to be in an engineering or arbitrary CRS.
It can be displayed alone or with other layers that are present, which
probably are in the same engineering CRS. However,
if other layers are added to such a view that are using a projection, then a
warning about mixing arbitrary or nongeoreferenced layers with georeferenced
layers will appear.
Legend Entries.
An important addition created by linking
to shapefiles via a shape object is gaining access to the styles used in the AVL
style file (*.avl) associated with the shapefile. The styles accompanying the
shapefile in its *.avl file were first used in V6.9
to style its elements and have been expanded to support their use in the
LegendView especially via network linked files.
Now these styles are used to create legend entries in the LegendView,
which appear and behave just the same as those for an internal vector object.
Linked shapefiles also now use the
built-in advanced TNT styling
procedures and these enhance the appearance and usefulness of the AVL legend
entries for that layer. A single
AVL style can be subdivided into new style entries, such as a population range.
Or, 25 colors might be assigned as styles in the AVL file to 200 national
boundary polygons creating 200 legend entries with the duplicate colors
repeated. The LegendView process
can detect this, make 25 color entries, and place multiple country names next to
each color. Both of these and other
related improvements applied via a linked shapefile and its associated ArcView
legend are illustrated in the accompanying color plate entitled Direct
Display of Shapefiles/Legends/Styles.
Alas, the needed AVL style file may
be missing when the shapefile has been created particularly outside the ESRI
Products and where the third party product is using the shapefile merely as a
convenient format in which to deliver graphical data.
If the shapefile is provided without its AVL styles, then TNT
styles can be assigned to the elements.
| Note:
Many shapefiles are being created to hold graphical geodata and associated
attributes but are not provided with associated styles in an AVL style file. |
DataTips and Multiple Tables.
Shapefiles have only a one-to-one linkage
between the elements and the attributes. A
TNT shape object has an implied
one-to-one linkage permitting an element to derive information directly from
more than one attached table. The
shape object link to a shapefile is used to define and contain any modifications
of this type, such as creating additional tables for the shapefile elements.
This permits the advanced features used for database attributes in TNTmips
to be used with these layers without altering the original shapefile.
These include adding additional tables, creating virtual tables, creating
single and multiple line DataTips, and other unique TNT
features for this linked shape layer as illustrated in the accompanying color
plate entitled Direct Display of
Shapefiles/Legends/Styles.
Now, when a shapefile is linked to and used in a TNTmips
view, it can be set up to provide the kinds of features you expect to have
available.
Unit Conversion.
If a numeric value is needed in multiple
units in a shapefile, it has one field for each converted unit.
TNTmips provides more than 50
units and their conversion. Identifying
the units of a single field in a shapefile permits a DataTip or multi-line
DataTip to display these values in any other units for features in the layer
even using virtual fields.
Frame Color.
The default color for the background for
your DataTips can be set using Support / Setup / Preferences.
Select a color to make them more visible for their associated view layer.
No more pale yellow DataTips on yellow or faded white map backgrounds!
Raster Display.
Automatically use the binary “null
mask” subobject if it exists. This
subobject is created by raster mosaic, reprojection, and extraction when a
single null value is not specified or is ambiguous due to inputs having multiple
or no null values.
Vector Display.
When adding a vector layer, a “density
check” is now performed to determine if it may be inadvisable to display the
vector at full view. If the vector
is determined to be too dense, a verification window is displayed with options
to not add the layer, add as normal, set maximum visible scale, and initially
hide the layer. The default maximum
visible scale is set automatically based on the density.
This setting helps avoid displaying the vector when the time required may
be long or the features may not be visually separated.
This new feature helps you avoid long delays and its use is illustated in
the accompanying color plate entitled Managing Display of Large Vectors.
* CAD Object Display.
Layer Controls.
The CAD Layer Controls window for the
display of CAD objects has been redesigned to use tabbed panels to support the
many new options for rendering CAD elements.
This window is now modeless, meaning that all you need to do at any time
to see these changes in the view is to press the “Apply” button in the lower
right corner and/or click on the Redraw icon in the view window without closing
the CAD Layer Controls. The window
has three tabs: Object, Elements, and Label.
Object Panel.
The Object panel contains the object,
style, and georeference selection controls, but adds the “Warp to Model” and
“Scale Range Visible” controls that were available in the V6.9
Vector Layer Controls.
Elements Panel.
The Elements panel consists of the
“Select” and “Style” controls and now provides the DataTip controls for
easier access. In V6.9
the DataTip settings were made via a Tools icon menu selection.
Label Panel.
Options.
The Label panel contains all the new controls for new CAD text styling
and rendering capabilities added in RV7.0.
The new option to “Clip Elements to Label Boundaries” operates in the
same way as the vector control of the same name available in V6.9
and causes the clipping out of a section of all the CAD elements around each
label. For example, lines are
clipped around, and do not run through labels.
The “Show Label Baselines” option will show the CAD text element
baselines used to position every label for rendering.
Frames and Leaders.
In RV7.0 the Frame Style
button brings up a tabbed window to enter label frame and leader line
parameters. This is the same panel
as provided in V6.9 for designing
vector labels and is also illustrated in the accompanying color plate entitled Enhanced
Sketch Annotation. Using the
Frame tabbed panel, the labels’ frame shape and margins; border style, color,
and width; and fill style, color, pattern and transparency can be designed.
Using the Leader Lines panel the labels’ leader style can be set as a
line with a width and color or as a triangular leader whose properties (for
example, line style, background color, and transparency) match those of the
frame it will extend out from.
During import or from some other TNT
activity, each CAD text element could have its own individual label frame style
assigned. Thus if “Style By
Element” is selected, the label frame styles set up using this control will
not be used in favor of the individual element frame styles.
GeoToolbox.
The cross section generation tool now
optionally generates a manifold object for the cross section drawn on a vector
object. This is discussed in detail
in the section below entitled Manifolds.
Annotating a Sketch Layer.
All the TNT
products including the FREE TNTatlas
and except TNTsim3D permit you to
create a sketch layer in a view that can be saved as a CAD object.
You can create attribute records in a table as you add features to the
sketch. You can set up this table
before or when you decide to draw a new element and create new records to go
with the elements. A sketch layer
is a quick way to sketch in data over an image in a field setting using a FREE TNTatlas
or to create a graphical annotation layer to add to the group for the current
view, perhaps for use in a layout.
Labels are added to your sketch layer
using text tools provided in the GeoToolbox.
RV7.0 permits you to improve
the appearance of labels used for annotation purposes by adding frames, frame
background color, and leader lines to labels.
These labels can have all the cosmetic properties previously introduced
for labels, such as rounded frame corners, transparent backgrounds, and
triangular leader lines. A single
label can also have multiple leader lines to point to several common features.
The accompanying color plate entitled Enhanced Sketch Annotation
illustrates the use of these new sketch features for annotating an image.
Profile View Tool Updates.
A new mode was introduced in the Profile
View tool. For a given set of
points it reads values from a mutiple raster set (hyperspectral set, time
series, and so on) and generates a profile where the x axis is the raster ID
number and the y axis presents the values from the raster cells.
This new profile is illustrated in the accompanying color plate entitled Graph
Values from Multiple Rasters by Cell Location.
Styling DataTips.
Background.
DataTips continue to be a powerful
feature that is relatively unique to the TNT
products and are easy to set up while producing more attractive results.
Conceptually DataTips provide value and/or text metadata about a specific
feature without cluttering up the geographical features in the view.
In V6.9 the standard
background, border, and text colors of DataTips and ToolTips can be uniformly
defined for all pop-in features via Support / Setup / Preferences / Interface
and use the ToolTip Colors button (now the ToolTip and DataTip Colors button) to
open the Color Editor window for this purpose.
Remember that after changing any of these from the default and selecting
OK, you must exit and restart any open process for the changes to take effect.
Styling Text and Values.
RV7.0
permits the DataTip contents to be much more attractive in form by the addition
of the use of text formatting control codes.
The text strings and value or virtual value of each line in the DataTip
can now be styled by including the TNT
text style codes in these lines. All
the {~} TNT text formatting control
codes can now be used to attractively structure the appearance of the DataTip.
These include defining the text style and font, color, tabs, and many
others.
Dynamic Frame Backgrounds.
A special text code has been added to
permit you to set and dynamically change the color of the background in the
DataTip frame. Use {~BG=red} or
{~BG=90,10,10} to set the background color of the text string for the DataTip to
the color “red” or to R=90%, G=10%, and B=10%.
This format code can be stored with or derived and added to a DataTip via
a virtual (computed) field or added to the prefix for the DataTip in the Layer
Controls. Thus several {~BG} codes
might be inserted in a DataTip string from various real and virtual sources and
unlike text format codes, only one background color can be applied to all of the
DataTip. The first {~BG} code
encountered is the one that will be used. The
accompanying color plate entitled Setting DataTip Background Color
illustrates this added feature.
As noted above, a color background can be
set up for all the DataTips using the Support menu.
Subsequently a virtual field can insert a new {~BG} code to override and
change this default background color. The
first layer with a background color specified for the nearest element determines
the background color of the DataTip, so the background color may change
depending on which layers are included in the DataTip for the current cursor
location.
To support this expanded use of the
DataTip text string for format codes, the prefix for each DataTip has been
expanded from 31 to an unlimited number of characters.
The application of these codes is illustrated on the accompanying color
plate entitled Add Styling to DataTips.
All the current available text formatting control codes (except ~BG,
which was added after this plate was printed) are documented on the reverse of
this color plate.
Use with CAD Layers.
| Important:
CAD layers can now have DataTips with all the current features available for
use with vector layers! |
The new CAD Layer Controls window via the
Elements tabbed panel now presents the same DataTip setup controls as the Vector
Layer Controls window. Now CAD
layers can also have the new GraphTip and Display Control Script features
discussed below.
HelpTips.
Remember that DataTips can be set up in a
special form referred to as HelpTips. These
are multiple line instructions that are spatially sensitive and pop in depending
upon where the cursor is stopped for a moment on the view.
This is an automatic transient pop-in effect.
HelpTips were introduced in V6.0
of the TNT products and can be seen
in the color plate entitled Data Logger APPLIDAT and now located at
www.microimages.com.
HelpTips are made up by attaching a table
with a long string field containing these instructions.
The vector is used to control the action of the cursor.
When the cursor is moved around the view and hovers, the instruction of
what to do next are a popin HelpTip, which is simply a text only, time delayed
DataTip. For example, the HelpTip
might pop in to instruct the user to “depress the left mouse button to select
…” while elsewhere in the same view it may present other instructions.
With the new format codes and enhancements, these HelpTips can now be
more attractively structured and presented.
GraphTips.
RV7.0
adds new, unique capabilities in the TNT
products to provide for interaction with the view, hidden layers in the view, or
with objects not in the view at all but that match some area of the view in
extent. The user selectable methods
of using Tool Scripts, Macro Scripts, Display Control Scripts, and others make
up one group of scripts all of which are implemented using the TNT
internal geospatial scripting language (SML).
The second group contains information that is set up in advance to occur
interactively. DataTips and
HelpTips were members of this group in V6.9.
RV7.0 adds new and powerful
capabilities of this type with the additional implementation of Enhanced
DataTips, GraphTips, and Display Control Scripts.
The editorial section above reviews how these features have evolved in TNTmips
by our joint efforts with you.
These new capabilities also grade into
each other and the terms DataTip and GraphTip are loosely applied to convey the
appearance of what these methods look like to the end user when they pop into
the view. The underlying code and
procedures for establishing these results is the earlier DataTip procedure that
has been extended to accommodate formatting codes and other new extensions.
Now the new Display Control Script is the basis for the pop-in display of
a GraphTip and more complex spatially determined GraphTip applications.
The Display Control Script is saved with the TNT
layout or group and a simple extension of the display process detects it and
runs the script to monitor the cursor movement and position.
If you add this layout with an associated Display Control Script to your TNTatlas,
it will be automatically run in that atlas. To assist you in reviewing some of
the properties of DataTips, GraphTips, and Tools (Tool Scripts) a color plate
entitled Use a
DataTip, GraphTip, or Tool? accompanies this MEMO.
Few alterations were required to RV7.0
to enable these various new powerful features. To get things working together
and tested, implementing a series of examples was needed.
The ability to explain these interactive samples in text is limited.
They are better illustrated by a series of sample applications.
The sample CD atlas of Afghanistan was prepared to permit you to try 4 of these concepts.
The sample DVD Property Viewer atlas for Lincoln, NE also illustrates Enhanced DataTips based on extensive property ownership
records.
Sample Display Control Scripts.
Each of the following sample Display
Control Scripts (DCS) can be downloaded from the TNT
script library at www.microimages.com/sml/.
Depending upon their size, all or a unique part of these sample Display
Control Scripts is printed on the reverse of the color plate that illustrates
it. These printed versions are
annotated to help you understand the script’s structure and procedures.
Local Time Zones.
This sample Display Control Script (DCS)
displays the current local time as a clock for any pause position of the cursor
on the view of the outline map of the world.
The color of the clock indicates if it is an acceptable time to call into
that time zone based upon a test of whether the current time there is within or
outside the programmed definition of local calling hours.
This sample script uses only a display of
the global political boundaries layer from the DVD entitled Global Reference
Geodata distributed with V6.9,
polygons of time zones from a hidden vector layer, and the time set on your
computer. You could dress up this
map by adding the world image layer to this display from this same DVD.
This GraphTip resulting from this DCS is illustrated in the accompanying
color plate entitled Local Time Zones.
Pie Chart and Bar Graph.
These sample Display Control Scripts (DCSs)
show two simple examples of how data relationships can be better illustrated in
a GraphTip than using text in a DataTip. One
presents a pie chart GraphTip of the family structure of the households in Nebraska
counties using TNTlite tutorial
geodata supplemented with population information.
This GraphTip is further enhanced by providing both the identification of
the sectors and the actual values of the sectors in matching colors.
The DCS simply draws what appears to be a text only DataTip below the pie
chart. The results of this DCS and
a discussion of its structure are presented in the accompanying color plate
entitled Pie Chart and Bar Graph. This
DCS in a slightly modified form can be interactively tested using the TNTatlas
of Afghanistan shipped with this RV7.0
of your TNT product.
Its operation in this TNTatlas
is also illustrated in the accompanying color plate entitled GraphTips in the
Afghanistan Atlas.
The bar graph GraphTip illustrates how a
DCS can provide another method to present interrelated data.
This bar graph GraphTip presents the age groups in Nebraska counties using TNTlite tutorial
sample datasets. Most of this DCS
is devoted to drawing these features. It
can serve as a model for your GraphTips that are oriented toward presenting
histograms. The results of this DCS
and a discussion of its structure are presented in the accompanying color plate
entitled Pie Chart and Bar Graph.
Enhanced DataTips and
GraphTips.
Similar results can be automatically
displayed quite differently in a DataTip versus a GraphTip depending upon your
objective. The accompanying color
plate entitled Enhanced DataTips and GraphTips illustrates this by
presenting the same information in both forms.
The DataTip is easily set up and formatted.
It presents slope and aspect numeric values in an attractive format for
the current cursor position on the color-infrared image in the view.
Both of these values are virtual fields computed from the cell’s center
at the cursor position. This
presentation is more than adequate for a forester studying the vegetation
distribution.
The corresponding GraphTip example uses a
DCS to compute and graphically present these same values as gadgets.
In this example, the cursor is being moved about on a geologic map
enhanced by the addition of shaded relief and a contour map. Making
these same slope and aspect computations in the DCS and presenting them in this
GraphTip form is oriented toward the way in which a geologist measures and
evaluates them. Also note that the
reference orientation for each of these computed properties is obvious in these
gadgets (which means, the baseline for the slope and the reference direction for
aspect). This information is
not obvious in the DataTip version but is easily deduced by the forester from
the image. This DCS in a slightly
modified form can be interactively tested using the TNTatlas of Afghanistan
shipped with this RV7.0 of your TNT
product. Its operation in this TNTatlas
is also illustrated in the accompanying color plate entitled GraphTips in the
Afghanistan Atlas.
Profile of Nearest Line.
For
any position in a view, a geospatial script can detect the nearest point, line,
or polygon within a specified radius. This
has been available for use in Tool Scripts and other SML
processes to find the closest element to the cursor position at which the mouse
is clicked. Now this can also be
used in a Display Control Script (DCS) to automatically detect the nearest
element of the designated type and layer. Thus
a DCS can automatically detect, access, and process the nearest element within
the specified radius when the cursor pauses.
The
accompanying color plate entitled Profile of Nearest Line illustrates
this and also that a GraphTip can actually be a graph.
This example is for a drainage network used as a vector object overlaid
on the display of a DEM raster object. Whenever
the cursor pauses near a drainage line, it is detected and highlighted by the
DCS. The script then traces the
profile of the drainage line along the DEM.
It then presents a GraphTip that is the elevation profile of the
highlighted drainage segment. The
elevation of the drainage at the nearest point on it to the cursor position is
also indicated.
In
a second example, the drainage overlays a color orthophoto but the GraphTip is
still its elevation profile. In
this case the DCS points to and uses a DEM, which is a hidden layer in the view
or it can be simply a raster object in a known path location.
One of the “under-the-hood” features of this and other geospatial
script applications is that it can automatically take advantage of the powerful
transparent spatial data management aspects of the TNT
products. For example, an off-line
DEM raster object (in other words, its path only is specified by the script) not
added to the view can be of any map projection, cell size, and extent as long as
it covers the requested drainage segment or part of it.
The script will still be able to automatically trace the drainage across
it and present the profile. In fact
some other program or action could substitute a new off-line object of the same
name and geodata type and the GraphTip would still be presented.
Even with all these various automatic geospatial conversions, the
GraphTip still pops in within a fraction of a second.
This DCS in a slightly modified form can
be interactively tested using the TNTatlas of Afghanistan shipped with
this RV7.0 of your TNT
product. In this atlas the nearest
road segment is detected, and the GraphTip is the elevation profile of the road.
This would be a particularly useful means of quickly choosing a route for
rapid movement in this mountainous terrain. Its operation in this TNTatlas
is also illustrated in the accompanying color plate entitled GraphTips in the
Afghanistan Atlas.
Display Control Script Approach Versus
Tool Script.
To present a GraphTip, a Display Control
Script is automatically run for each pause of the cursor in the view.
A Tool Script is a tool selected by an icon from a number of tools/icons
on the icon toolbar of a TNT view.
Some of these tools/icons are connected to preprogrammed, built-in TNT
features (in other words, Zoom In, Select, View-in-View…).
However, if they are connected to a Tool Script, they run the associated
script. A common Tool Script
procedure in such a script is to await a mouse click in the view and then act on
that position in the view. Once a
pause is detected in a DCS or a click in a Tool Script, quite similar results
can be achieved since both then have access to and use the extensive
capabilities of the TNT geospatial
scripting language. However, the
DCS is automatic and pushes information at you, while the Tool Script requires
interaction and is used to pull information to you.
As a result, while these scripts can have results, they are different in
some aspects of their structures as illustrated by the following sample scripts,
which could produce an identical visual result.
Spyglass View Display Control Script.
This example illustrates that a Display
Control Script (DCS) can present GraphTips that display portions of any TNT
raster or geometric object. When
the cursor is paused in a view, this DCS detects the position of the cursor and
inserts a circular view of a specified TNT
layer or TNT object into that
circlular area. This GraphTip
closes when the cursor is moved and reappears at the next pause position.
It acts something like an automatic View-in-View.
The contents of the interior can match in
scale and features at the edge of the circle (as in these examples), be zoomed,
or whatever you define in the script. The
contents can be from a raster, vector, shape, or CAD layer or directly from an
object not used in the current view. When
an object is used, it does not have to be in the same Coordinate Reference
System, datum, cell size, scale, element size or length, and so on.
All these possible conversions are handled automatically in a geospatial
scripting language, such as SML.
The accompanying color plate entitled Spyglass
View illustrates a DCS that fetches a matching circular area from a color
image raster object and displays it in a view of a topographic map.
A second illustration on this plate shows a circular view fetched from a
soil map vector object. This
GraphTip contains the soil polygons matching the map scale and filled with
transparent colors so as to not obscure the underlying topographic map features.
Note also that each time this GraphTip moves, it may have a different
subset of the soil types within it. Thus,
the LegendView is also updated to identify only those soil types currently
exposed in the circular view.
Clearly a lot of activity is being
conducted by this kind of DCS and yet it is interactively able to push these
GraphTips into the view in a second from large objects.
This is because years of work have gone into the TNT
processes (pyramiding rasters, optimizing vectors, indexing tables, …) and,
thus, SML toward seeking speed in
accessing small areas from what can be potentially very large geodata objects.
Any shape and size of an area that can be defined or computed in this DCS
or the Tool Script that follows can be used instead of a circle.
Some examples would be to compute a region(s), a viewshed, a route,
buffer zone, and so on. A
geospatial scripting language is ideal for such purposes.
However, complex scripts may take seconds to complete and would be more
suitable for running only on demand in a Tool Script.
Pop-In View Tool Script.
A Display Control Script can present
dialogs but often does not as its purpose is to automatically present or push
information to you. Thus, a DCS is
usually designed around your setup of known layers and objects in known positions
such as in a TNTatlas.
A Tool Script runs on demand from an icon.
As a result the geodata, control parameters, and other inputs it uses can
be obtained at the time of its use by including XML defined dialogs in the
script. As a result Tool Scripts
can be designed to provide a widely used and reused tool since the location and
type of geodata each uses can be requested from any user.
Their operation can also be modified by the user, such as setting the
radius or picking a shape for the GraphTip in this example.
They can also be used to present forms, populate tables, and start other
programs, such as a browser with a URL containing the cursor’s position.
The accompanying color plated entitled Pop-In
View illustrates a Tool Script that also fetches and inserts matching
circular views from a raster object and from a vector object that are not part
of the current view. However, this
tool differs because when its icon is clicked, it requests the type, location,
and name of the object to insert into this circular area.
Once you have provided this input, the circular view of that object is
inserted at any position of the cursor when the mouse is clicked.
It remains in the view until a new position is clicked, another tool is
selected, or the view is refreshed. Since
this is a Tool Script and requests all its needed inputs, it can be added to the
icon tool bar and reused with any TNT
desktop product.
The color plate illustrates two results.
In one the object selected for the interior of the Pop-In View is a
raster object that contains a combination of a vector object of a geologic map
overlaying a shaded relief raster object. The
image in the view is a color orthophoto. The
second example uses the 1-foot color orthophoto raster in the Lincoln Property
Viewer TNTatlas in the view.
The Tool Script then inserts a circular view of the vector object for the
floodplain polygons from the floodplains vector object in this same atlas.
These polygons are found, clipped, and inserted into the circular area
with transparent colors so that the image features show through. As
noted above, you can download this Tool Script from www.microimages.com/downloads/scripts.htm
and add it to your icon tool bar using the instructions on page 47 of the
tutorial entitled Writing Scripts with SML.
You will also find in this tutorial 14 pages of information about other
Tool Scipts, Tool Script templates, and how you can create you own tools.
Miscellaneous.
The default mode for DataTip viewing is
set to “all layers” if you have not previously set it to something else.
An
option has been added to not draw labels until other layers in the group have
been drawn.
When
the scale range for viewing a layer has been set, it is added to the DataTip
that pops in when you hover over the layer name in the LegendView.
You can now set which raster is to be
used when a 1X zoom is requested. This
is set on the Options panel of the Raster Layer Controls.
It can be used to control which raster to use in situations where rasters
of significantly different cell sizes are mixed.
A new zoom mode icon (red magnifier with
a “+” sign) is now provided by default on the icon bar of the view window.
Selecting this icon puts the cursor in the mode to continually zoom in 2X
at the feature you point to when you click the mouse. If
you click the numeric keys 1 thru 4 in this mode the zoom will be at the cursor
and have the same result as if you were in any other mode.
For example, if you type “1” the zoom will be to full resolution and
centered on the cursor.
You can now highlight and show an element
from a hidden layer.
When a view has any hidden layers/groups
then by default these are hidden in the View-in-View tool.
* 3D Display.
Surface
Rendering Modes.
The
rewrite of the TNT 3D rendering
engine and previously available features is complete in RV7.0.
Dense Ray Casting, Variable Triangulation, and Sparse Triangulation
surface rendering options have been fully implemented.
These rendering modes (selected from the Surface Layer Controls dialog)
determine the elevation accuracy and speed of rendering of the 3D surface model
on which your other geospatial layers are draped.
The two original surface rendering methods of Ray Casting and
Triangulation have been removed from RV7.0
TNT products as their older code and
further application serve no useful purpose.
Now all their special features such as pedestals, stereo viewing, …
have been implemented and improved for the 3 new surface rendering options.
The
accompanying color plate entitled 3D Surface Rendering Modes compares
these 3 modes and discusses their speed versus quality differences.
The conceptual design of these surface rendering modes was introduced in
this same section in the Release MEMO for V6.8.
All three surface rendering options also can be used in combination with
any of the advanced texture filters described in the Release MEMO for V6.8.
Now that surface-rendering and texture-filtering options have been
revamped, effort will be directed toward speeding up these rendering methods to
enable more interaction between concurrent 2D and 3D views. Another possible
area of improvement in this new rendering engine would be to preserve more of
the vector nature and attributes for overlaying geometric objects.
Displaying
non-horizontal, planar or curvilinear structures called manifolds has been added
and is discussed in detail in the section below entitled Manifolds. Since
the display of manifolds is in a TNT
3D view, it does not use DirectX or OpenGL, but internal MicroImages code.
This means that manifold viewing is cross-platform and automatically
incorporates many important 3D view features, such as the ability to render a
high resolution version of the current view into a file for printing, extension
via SML, direct integration and
geographic correlations with the 2D views, and so on.
Faster Wireframes.
Moving
wireframes around is significantly faster because it now uses the appropriate
terrain pyramid tier to generate the wireframe.
Wireframe manipulation is commonly used for quick preview of this
skeletal form of the terrain to make adjustments to the viewpoint.
V6.9 always used the full
surface raster layer for this purpose requiring more data to be read in a
situation where it was not needed.
Faster Flat Surfaces.
Rendering
of multiple texture layers on a flat surface (which means, without a terrain
surface) is much faster. This is
actually how a vertically-oriented stereo view is rendered, as discussed and
illustrated in the Stereo section below.
In a stereo view parallax is introduced into the textures from the
terrain layer, but their time consuming depth rendering is now bypassed.
Transparent Layers.
The
transparency properties set for individual texture layers are now applied in all
surface rendering modes. When
multiple texture layers overlap in any given perspective view, their
transparency effects combine in the order in which the layers are rendered,
which is from the bottom of the layer stack (first layer added) to the top (last
layer added). This statement
applies to multiple texture layers assigned to a single terrain surface as well
as to overlapping textures from different vertically-separated terrain surfaces.
This
kind of transparency, called layer order transparency, was available in V6.9
only with the old Ray Casting surface rendering mode.
Now you have this capability available for use with all of the
high-quality surface rendering modes in your 3D views, both on-screen and in
printed layouts for data areas of any size.
For example, you can set a different transparency for each vector polygon
fill style and it is used when this layer is draped over an image.
Or, you can set the transparency level directly in the 3D controls for an
entire raster object, such as a scan of a geologic map, to visually merge the
map with an underlying opaque relief shading or panchromatic image layer.
Examples of the use of layer transparency are illustrated in the
accompanying color plate entitled Transparency and Relief Shading in 3D Views.
Remember that these transparencies are cumulatively applied bottom to top
as you view the layers in the current perspective.
Any opaque layer that is encountered will obscure the view of earlier
layers for that screen pixel.
Relief Shading.
Relief
shading can now be requested and manipulated directly while using any of the new
3D surface rendering methods. (In V6.9
relief shading in 3D only worked with certain of the older surface rendering
modes.) You can vary the sun
elevation and azimuth, as well as its intensity, via the vertical scaling of the
terrain layer. The relief-shaded
layer can then be used as a base texture layer and partially exposed by making
the 2nd and subsequent added texture layers, such as an image or map, partially
transparent. Adding this shaded
relief base enhances the visual clues in a 3D view and creates the illusion of a
sun position for all views. Of
course if the image texture layer has obvious feature shadows, this will dictate
the best angular position of your sun settings.
The accompanying color plate entitled Transparency and Relief Shading
in 3D Views illustrates various uses of relief shading.
Pedestals.
Adding
cosmetic pedestals to your 3D rendering has been improved and is available for
all 3 new surface rendering methods. A
pedestal for a texture layer is drawn vertically down from the outermost edge of
real data values in the texture or its underlying terrain.
A curving real data boundary, therefore, produces a pedestal with curving
walls. Pedestals now also provide
other advanced features not previously available, such as smoothed color, color
shading, and transparency.
Height.
A
pedestal is drawn vertically downward from the real data edge to the base
elevation specified for it in the Raster Layer Controls dialog.
This pedestal base elevation is a value you select relative to mean sea
level.
Color and Transparency.
The
color of the pedestal is selected from any color using the same dialog as above.
This color is automatically shaded according to the current position of
the sun, which can be set using the 3D Viewpoint Controls dialog.
You are even able to set a transparency level for the pedestal color.
The use of this feature is particularly effective for upward pedestals
called fences discussed below. However,
pedestals with a high transparency setting might also be used to put a frame
around a 3D view with multiple, but offset, terrain layers.
Irregular Terrain Boundaries.
Almost
all published illustrations of 3D views with pedestals show views that have an
outer boundary in the shape of a parallelogram or trapezoid because the
rendering process shows the full rectangular extents of the geodata.
In the TNT products, your
single 3D view can be irregularly bounded and use stacked or edge matched
terrains that are irregular in shape. If
your terrain and/or texture edge is irregular, then the pedestal follows that
shape. This is illustrated in the
accompanying color plate entitled Pedestal and Fence in 3D Views.
You are no longer restricted, as previously and in many competitive
products, to using a pedestal only around a rectangular data area.
If your terrain(s) or texure(s) have holes in them, then appropriate
pedestal sections can be viewed through these holes as a function of the view
angle.
When
pedestals are curved, they are attractively shaded as illustrated in this same
color plate. Pedestal shading
requires a complex process of smoothing the edge of the data area that generates
the pedestal shape, since this actual edge is always defined by raster cell
boundaries and so is not inherently smooth but saw-toothed.
There is also the issue of deciding when an inflection in the curved edge
of the data is a real abrupt inflection in edge that should be observed and
appropriately shaded versus being only meaningless spatial noise along what
should be interpreted as a smooth edge. These
subtle but important considerations are why irregular pedestal boundaries are
not supported in other products. However,
if not overworked, TNT produces very
attractive shaded pedestals for your irregular area.
You can mask out marginal surface areas that have distracting or ugly
features, such as an existing development, and these areas are then not rendered
in the view. The pedestal
surrounding the remaining real data area focuses attention on the site of
specific interest.
Fences.
This
new 3D view embellishment is simply a pedestal drawn upward.
Since a pedestal and fence can be applied simultaneously to the same
texture in the view, these features have their own independent sets of identical
controls for height above base, color, and transparency.
The meaning and operation of these controls is discussed above.
An
effective use of a fence is illustrated in the color plate entitled Pedestal
and Fence in 3D Views. In the
illustration, the pedestal is rendered downward and opaque to provide a base and
frame for the rendering. It also
emphasizes the 3D depth with its irregular terrain intersection with the texture
along the front edge. However, even
in this small print it can be seen that the upward highly transparent pale blue
fence adds a subtle corner effect further creating the illusion of depth.
Fences with high transparency can create this corner effect or with low
transparency build a sandbox effect around the terrain.
Stereo.
Rendering
a 3D perspective view as a stereo pair is now supported for all 3 new surface
rendering methods. Once a suitable
3D perspective view has been created, it can now be viewed in high quality
stereo using a stereo viewing device such as a 3D monitor or a mirror
stereoscope. Simply select one of
the 4 available stereo display modes. You
can include any of the new 3D viewing features in your stereo view including the
new manifold surfaces, advanced mipmap texture filtering, and all the other
features introduced incrementally in the previous TNT
releases of this 3D viewing process. If
you wish to create a straight down view in stereo, simply adjust the viewpoint
controls to place the viewer above the terrain with a 90-degree pitch.
The
TNT products can render the stereo
view in any of 4 stereo modes to match your choice of stereo viewing device.
To select the stereo mode, open the Stereo Viewing Options dialog by
choosing Stereo Setup from the Perspective View window's Options menu.
These various rendering methods are illustrated in the accompanying color
plate entitled Stereo Viewing Modes.
Almost every common stereo viewing device available uses one of these 4
display methods. Recommended
viewing hardware for use with these methods is discussed below.
| Note:
2D and 3D views can be switched into stereo mode in the FREE TNTatlas
and TNTlite, the new low-cost TNTview,
and TNTedit and TNTmips.
|
Separate Frames.
The
Separate Frames mode is new to the TNT
products with RV7.0 and displays
separate side-by-side images with parallax added.
This is the mode that provides very stable stereo viewing specifically
for use with a mirror stereoscope. Dual
high quality images can be displayed side by side on a 21" or a larger flat
monitor or a pair of flat monitors and a stereoscope set up over them.
To adjust the placement of the side-by-side views to fit your
stereoscope, measure the distance between its optical axes and enter this value
in the Stereo Viewing Options dialog. Since
the use of dual monitors introduces a bezel area between the 2 views, the width
of this obstruction also can be entered in this dialog to further adjust the
frame separation.
Interlaced Columns.
Interlaced
Columns mode creates images with appropriate parallax in a single view.
One image is displayed in the odd-numbered columns and the other in the
even-numbered columns. This mode,
depending upon the viewing device, reduces the horizontal resolution by a factor
of 2. This mode is used by
direct-view stereo monitors (no glasses, shutters, …) that use a special
screen surface to optically deflect the odd-numbered columns to one eye and the
evens to the other.
Interlaced Lines.
Interlaced
Lines mode creates two images that are spliced in the opposite direction from
column-interlaced mode. Image
lines from one image are displayed in odd-numbered screen lines and from the
other in even-numbered lines. Parallax
is introduced by slightly offsetting the starting positions of the interlaced
lines. On older interlaced display
systems, shuttered eyeglasses are synced with the display to force each eye to
separately view each of these images. Another
approach uses a shuttered screen and polarized glasses to create stereo.
Both of these methods, even in their latest incarnations, can quickly
introduce eye strain.
Anaglyph.
Anaglyph
support is included because of its simplicity and the ease with which the
viewing device can be acquired. Its
biggest advantage is that stereo images can be cheaply printed and distributed
in any quantity with a free pair of glasses.
There are no special settings required in using this option.
In the Stereo Viewing Options dialog, it’s simply a choose it and use
it option. A color plate entitled Anaglyph
Stereo Viewing accompanies this MEMO to compare this method of viewing with
those that follow and require special equipment.
Stereo Equipment Recommendations.
Stereo
viewing on a monitor has been an entertainment gimmick except when required for
special purposes such as soft photogrammetry. If good quality stereo
viewing was available, we would all be using it for TV, for games, and our
geospatial visualization. Low-cost
viewing devices produce flicker and quickly give you a headache. Better quality
devices have cost too much for general purpose use. For these reasons,
minimal previous effort has been expended in this aspect of visualization in the
TNT products.
However, now some approaches do produce acceptable results. Hardware
devices are becoming available that make it possible to view stereo for longer
periods without flicker and at acceptable additional cost.
This hardware and corresponding stereo modes in 3D views may now make
stereo viewing of increased interest to you.
Suggestions for more photo-like TNT
stereo stations are discussed and illustrated below.
For more information on stereo add-on devices and monitors please see
www.stereo3d.com.
Stereoscopes.
The
analog, tube monitors of the past were commonly multisync/multiresolution with a
continuous phosphor coating and had a tendency to jitter at the pixel level.
They were also large and bulky adding to the difficulty of using them in
pairs with a stereoscope. Now,
digital flat panel monitors display images at a single resolution to specific
discrete cells. This pixel
stability enhances their use to present a pair of photo-like stereo images for
viewing with a mirror stereoscope. A
single, large monitor can be used or a pair of smaller units. The
stereoscope has been around almost 100 years as a low-cost means of viewing
pairs of images in stereo with minimum eye fatigue.
Using these devices can produce near photo quality stereo, which can be
viewed longer than other approaches with a minimum of eye fatigue.
Inexpensive Mirror Stereo Station.
The accompanying color plate entitled Inexpensive Stereoscope Viewing
shows the lowest cost mirror stereoscope design.
It uses twin 15" or 17" flat panels, 1024 by 768 pixel monitors
you may already be using for your TNT
products. These can be positioned side by side laying flat or on an angle with
their bezels in contact in the middle. The
plastic mirror stereoscope (called a GeoScope™) can be positioned across them
and the stereo view mode selected in 3D display.
As noted above, the separation of the stereo frames can be set by
entering the distance between the optical axes of your stereoscope, and can be
automatically adjusted for the dual-monitor setup by entering the width of the 2
monitor bezels. These values need
to be set only once for this setup. Since
you are already using dual monitors with your TNT
product this is a US$200 approach.
Mirror Stereo Station.
The accompanying color plate entitled Stereoscope Viewing shows
more expensive, higher quality mirror stereoscope design.
It uses a single, 23" pan-wide flat panel monitor of 1920 x 1200
pixels yielding 960 by 1200 pixels for your stereo view.
It is available from Apple, Sony, and HP in an identical form factor and
price and is also an excellent choice for the general operation of your TNT
products if you like to use only 1 monitor.
A popular high quality mirror stereoscope with first surface mirrors and
quality lens fits almost perfectly across this single monitor over your
photo-like stereo images. You only
need to set the optical axes default separation.
This kind of monitor is US$2000 and getting cheaper while the stereoscope
is less than US$800.
Direct View Monitors.
Flat panel monitors are now available
that directly display stereo images without any glasses or other secondary
media. These monitors display dual
images that are column interleaved. These
alternating columns of pixels are then deflected odd-numbered columns to the
left and even-numbered columns to the right by a special surface on the screen.
Each of your eyes then sees only one of these color images.
With a high frame rate the two parallax-shifted images merge into a 3D
image projecting out from the monitor’s surface.
Using alternating images for stereo cuts in half the horizontal
resolution of this monitor. However,
at any time you can physically shift these monitors from this stereo behavior to
flat mode and use them just like any other monitor for 2D work at full
resolution. At this time, none of
MicroImages’ staff have first hand direct experience with the quality of the
stereo produced on these monitors.
Sharp Direct View 15" 3D
Monitor.
In 2003 Sharp released a portable (model
Actius RD3D) with a 15" flat panel with column interleaved stereo using
their parallax barrier technology for direct viewing at about US$3000.
Late this year they began shipping a stand alone 15" flat panel
monitor (model LL-151-3D) using their direct view stereo technology for US$1500.
This is a 1024 by 768 resolution monitor that can be switched between
normal 2D and 3D display modes in software or with a physical button. The
3D mode repeats the columns in the view for each eye.
This halves the resolution from 1024 to 512 by 768.
This monitor is only now becoming readily available.
MicroImages clients who have used it for viewing TNT
manifolds and surfaces in 3D report it is very good.
MicroImages has taken delivery on this
monitor and the unaided stereo viewing is very sharp and clear.
If the monitor is switched out of stereo to full resolution 2D viewing,
you can not detect that this is a stereo monitor.
The only drawback is that the viewing position is very narrow. However,
you can move your head closer or father from the screen and maintain stereo as
long as your head remains in this narrow viewing angle.
This is particularly important to those of you who wear glasses, and
especially those with bifocals and trifocals.
Holding the head still for playing games on it would be difficult.
For professional applications where the view or simulation is going to be
studied, this is more of an annoyance when its crystal clear stereo quality is
compared to any other stereo viewing device. For
example, everyone is familiar with viewing images with a stereoscope that
requires a very specific position for your head!
More information about this monitor can be found at www.sharp3d.com and
it has many reviews on the Internet.
Larger
Direct View 3D Monitors from SeeReal.
TNT
stereo views of manifolds, surfaces, and other 3D results were recently
demonstrated at a show to geologists using the larger but much more expensive
SeeReal monitors manufactured in Germany. These monitors use a
different
technology to route the images to each eye than Sharp’s monitor but use the
same TNT column interlaced mode.
They have several products in this line and their top model is the C-nt
at about US$10,000 (current prices are not posted anywhere on the Internet).
It is a 20-inch flat panel with a 1600 by 1200 pixel resolution.
Again, in stereo the horizontal is halved to 800 by 1200.
More information about these direct view stereo monitors can be found at
www.seereal.com/EN/products_cnt.en.htm.
Non-Topographic Stereo Applications.
Since the costs and quality of stereo are
now in an acceptable range (US$200 to US$3000), your potential uses of stereo
should not be limited to obvious conventional applications such as viewing
images, geologic and topographic maps, and manifolds, all of which require a
DEM. There are more exotic forms of
stereo and 3D viewing in the TNT
products that do not require a DEM or vertical manifold georeferencing. The
following are some sample ideas. Consider
how you can extend these suggestions for non-DEM related stereo applications to
the interpretation of your project’s research materials and objectives.
It is also important to realize that stereo applications are not
restricted to georeferenced geodata. Your
rasters and graphical datasets need only be internally registered, such as
derived from the same source rasters, or in the same engineering or Cartesian
coordinate framework to be used in stereo in the TNT
products as outlined below.
Vegetation Stereo.
A common stereo application is to compute
the green canopy biomass raster object from a color infrared airphoto or
satellite image using a vegetation index equation (for example, NDVI and many
others). This new raster will
register exactly with the original multispectral images.
It can then be chosen as a pseudo-DEM and the natural color or color
infrared renditions of the original images draped over it in 3D and stereo.
If the imagery is a color infrared airphoto or 1-meter synthetic
satellite image of a crop field, the variations in its canopy biomass within
each crop field will be dramatic. If
the color image is derived from a 15-meter Landsat, then the variations in
forest canopy density and conifer (needle) versus deciduous (broad leaf) species
will be enhanced in stereo and perspective.
Image Ratios.
Geologists frequently interpret
multispectral images by using special color band ratios selected to enhance the
surface distribution of a group of geologically related materials. From these, a
rough map is sketched of the distribution of materials in that enhancement.
Then another enhancement and its grouping of materials is mapped and so
on. As this procedure moves along,
the inter-relationships between these groups is worked out by going back and
forth between the ratio images and overlaying the current composite surface map
and its subsurface extrapolations.
Many image processing steps may be used
to enhance these ratio or other image presentations to aid in this
interpretation such as vegetation removal, corrections for terrain induced
radiance and atmospheric effects, some level of auto interpretation,
orthorectification, and so on. However,
in the end it will be the expert geologist that interprets these rasters and
builds the electronic or physical map of the geology of the area.
Geologic interpretation and its
refinement gets an obvious improvement when a DEM is used to present these
enhanced image products in stereo and perspective, along with subsurface
profiles as manifolds, and their current interpretations as overlays.
The initial sketching, revision, and mapping activities can then be
carried on concurrently in a separate 2D map view.
However, these geologic interpretations can also be improved if stereo
and 3D perspective viewing are used to enhance the geologic characteristics of
the area rather than the topographic relationships. The following is an example
of how stereo can be used by deliberately omitting the DEM and using stereo with
just the enhanced imagery to improve the interpretation of the
inter-relationships between groups of geologically associated surface materials.
A color enhanced ratio raster that has
been optimized to help map a group of related surface materials (class A) can be
displayed in stereo and 3D using as a “pseudo-DEM” a second ratio raster
optimized for another group of surface materials (class B).
Relationships between these groups will now be enhanced in stereo and
interpretations made taking these stereo effects into account (for example,
class A can be here but that means class B can not be).
In TNTmips it is then easy to
toggle a third or new ratio raster in as the “pseudo DEM” and compare in
stereo a new combination of surface groupings (class A over C) to refine the
interpretation of the location of the group of surface materials in the color
raster (ratio class A) before drawing any lines.
These procedures are easily applied to images from multispectral image
sources such as Landsat or ASTER. Viewing
combinations of more advanced automatic interpretations derived from
hyperspectral images in this way can create many possibilities of this type
where final material identification can not be completed or trusted entirely to
the automated processing, such as in a military decision.
Water Quality.
Subtle variations in water features, such
as the boundaries between algae concentrations and pollution levels, are often
hard to quantify. Often this is
improved by enhancement of multispectral images or the addition of thermal
images. If visual interpretation
and mapping are used, these kinds of enhancements can be combined in stereo as
above to improve their interpretation.
Stereo Plans.
Tentative plans include adding manifolds
to TNTsim3D since DirectX and OpenGL
can be readily used to introduce stereo viewing into this real time simulation.
Stereo support can then be easily added to use this in TNTsim3D.
This combination will permit you to move the viewpoint of a stereo view
around in real time to help you understand the spatial relationships in a
simulation using 3D terrain surfaces and/or manifolds.
Communicate what you need next in stereo visualization to get it on the
development list.
©MicroImages, Inc. 2010 Published in the United States of America
11th Floor - Sharp Tower, 206 South 13th Street, Lincoln NE 68508-2010 USA
Business & Sales: (402)477-9554 Support: (402)477-9562 Fax: (402)477-9559
Business info@microimages.com
Support support@microimages.com
Web webmaster@microimages.com
| 