The SEDRIS Data Representation Model
APPENDIX A - Classes
Reference Vector
|
|---|
Class Name: Reference Vector
Subclasses
This DRM class is concrete and has no subclasses.
Definition
An instance of this DRM class specifies a unit vector that is used to
specify a direction, normal, or axis in the manner specified by the
vector_type field. It
serves as a unit vector in those spatial reference frames with
compatible vector space structure. In GD, GM,
GEI, GSE, GSM, and SM coordinate systems, it is a vector in the canonical
LTP at the location of the <Reference Vector>
<Location> component.
Primary Page in DRM Diagram:
Secondary Pages in DRM Diagram:
Example
- A <Reference Vector> contained by a
<Polygon>, representing its geometric normal vector.
This would have a vector_type of SE_FACE_NORMAL.
- A <Reference Vector> contained by a
<Polygon>, representing a normal vector that is used
for rendering purposes, i.e. to calculate colour and shading when rendering
the <Polygon>. This
<Reference Vector> would have a vector_type of
SE_REF_VEC_TYP_RENDERING_NORMAL.
- A fence modeled as a polygon has radar cross sections that are dependent
on aspect angles (azimuth and elevation). These aspect angles are defined
with respect to the polygon's normal and its azimuth
<Reference Vector>. This
<Polygon> (fence) has two
<Reference Vectors> (a geometric-normal and
an azimuth-reference). The geometric normal is the unit vector
perpendicular to the <Polygon> pointing away from it
on its outside face and the azimuth reference vector points straight up
and is in the plane of the <Polygon>.
- A segment of the road has a reflector (actually, a retro-reflector) on it
and is modeled as a <Line>. The
<Line> has a normal vector that is
perpendicular to it and an azimuth reference parallel to it. This is
sufficient to describe radar cross sections of the road as a function of
aspect angles. However, the normal vector for the infrared bands depends
on the orientation of the retro-reflector, not the road. This because
radars see the road but IR (or more obviously, car lights) see the retro-
reflector. In this example, the <Line> has four
<Reference Vectors>
(radar-normal, radar-azimuth, ir-normal, and ir-azimuth).
- A normal vector used for reflectivity/emissivity calculations.
This would have a vector_type of SE_REFLECTIVITY_EMISSIVITY_NORMAL.
- A vector specifying the direction an <Infinite
Light> illuminates. This would have a vector_type of
SE_LIGHT_DIRECTION.
FAQs
- Why so many different types of <Reference
Vectors>? Aren't all or many of these the same unit vector with
different names?
- These vectors coincide with each other in many cases, but generally, each
each can be unique. For a flat large window, as an example, the geometric
normal will coincide with the radar-normal and also the ir-normal, but for
complex objects such as an aircraft, geometric-normal is not applicable
and radar- and ir-normal does not coincide.
- Why can a <Base Vertex> have only one
<Reference Vector> when a
<Polygon>, <Point>, or
<Line> can have more than one?
- A <Base Vertex> will not require more than one
<Reference Vector>, but a
<Polygon> may need more than one to represent a
combination of geometric and/or sensor-related vectors at the same time.
In other words, a side of the building represented as a
<Polygon>, a ship as a <Point>,
or a bridge as a <Line> require at least two
<Reference Vectors> defined, normal and
azimuth, to fully describe aspect dependent characteristics of the
physical object represented. Vertices, on the other hand, do not represent
these types of physical objects and therefore do not require multiple
<Reference Vectors>. A vertex can have one
rendering-normal as used for smooth shading, but has no use for aggregating
any other types of <Reference Vectors>.
- Why does <Reference Vector> have an
optional <Location> component?
- To support the ability of the SEDRIS API to automatically convert
<Reference Vectors> between spatial reference
frames, including non-vector space coordinates, such as Geodetic.
- If a <Location> component is required for some
conversions, why is it an optional component?
- In most cases, a <Reference Vector> will
inherit an appropriate <Location> component. In those
cases where an appropriate <Location> cannot be
inherited, a <Location> component is required by
constraints.
See Required
Reference Vector Location.
Constraints
Composed of (two-way)(inherited)
Component of (two-way)
Prev: Reference Surface.
Next: Reference Vector Control Link.
Up:Index.
|
Last updated: October 1, 2002
|
Copyright © 2002 SEDRIS™
|
|