|
Return
to March 2003 Newsletter
True 3D Mapping Reveals the Inaccuracy of 'Symbol'
Mapping
Part III by Steve McKinzie
 In
Part 1 and 2 (Forensic
Mapping Challenge, Feb 2003, vol V, issue 2) we discussed
the identification, documentation and validation of an individual
point measurement. The measured points are used to create a database
of measurement information from which to create the map. The database
is manageable and protected. Remember, to each position the rod
person has assigned a code. The code sometimes looks akin to a
foreign language. It is really nothing more than an abbreviation
for a graphic attribute. The coding can be an assignment for a
straight line, curved line, (concave or convex) arc or symbol.
Most software suppliers provide a default coding library to launch
your endeavors and accompanying editor. However not all software
is based on the geometry to accurately recreate the scene.
Enough
emphasis cannot be placed on the responsibility of the rod person.
Part of which is the assignment of the forensic code to the measured
point. The rod person must be able to visually detect the differences
between a straight vs. curved line, crowns, grades, and vertical
curves in the area to be documented as well as classification
and characterization of physical evidence.
In
order to map a straight line for example, two points must be measured.
Each Point is assigned a forensic code and an instruction to draw
a straight or curved line. Traditionally, you use a command code
with the graphic instruction to start and stop the line. A code,
for example ZEP1, is an instruction to begin a straight line using
the graphic instructions found in the library under Edge of Pavement.
In this example EP1 is a black line 1 pixel wide.
Figure 9
| Point
1 starts at position 0,0,0 (0 north, 0 east, 0 elevation)
and ends at position 5,0,0, XYZ if you prefer. It is important
to understand the measured position may not always reflect
the line position. If for example point 2 is given a Z position
of positive 5 or five units above it's origin, the horizontal
distance is different than the slope distance. In a topographic
or plane view no change can be detected. However if we examine
the line and data from an ISO view we see the line is actually
much different. |
Figure 10 Fig 9 Line Data
|
Figure 11 Point 2 w/Elevation |
Figure 13 ISO View of Fig 11 |
Figure 12 Fig 11 Point Data |
In this demonstration
the importance of true 3D mapping and hard supporting data is
realized. A complex curve may have multiple measured points to
accurately identify the curvature. Each group of multiple points
will have the scene graphic code, however a different command
may be found in order to facilitate the line curvature between
convex and concave.
Figure 14 Topographic of Complex Curve
In
this 2D topographic view, (Figure 13) the line curves fix the
model. When we look at the same line from a profile view the importance
of elevation data is appreciated.
Figure 15 Blue 2D; Red 3D
Yet
another example is to create a curve or arc from three measured
points, see Figure 16. The difficulty in mapping such an arc is
finding a constructed true radius.
Figure 16 Three Point Arc
Far
too much importance is placed on the width of lines. In nearly
all cases a line is nothing more than a boundary. A line depicting
the edge of pavement adjacent to a gravel shoulder is not a perfect
edge. We use a perfect edge in a drawing to depict the boundary
- not the edge. A fog line near the pavement edge is usually four
or six inches wide. The tighter the budget, the narrower the line,
but that's another story. In our case we are depicting the boundary
by mapping the center of the line. The line width can be accommodated
when preparing your map as a demonstrative exhibit. Otherwise,
save your time and energy for accuracy and foundation.
Our
imaginary map is now recorded point by point. We have meticulously
assigned codes to each point, thereby controlling our graphics.
When the map data is processed the geometry is controlled by measured
data and the lines, arcs, and curves between those points are
controlled by the graphic code assigned to each point. This map
is technically correct.
Not
all landmarks found at a site to be measured are conducive to
identifying with a line. While it could be done, it is far too
time consuming. Remember, speed is what you promised the boss.
Let's take the examples of a fire hydrant and a street light.
Both items are regularly found at crash scenes, just as a shell
casing can routinely be found at a violent crime scene. The entity
may be placed in the map for visual reference. If we are dealing
with a two vehicle intersection collision and the fire hydrant
present was not struck and the crash occurred at noon, do they
need to be in the map? For the technical analysis or reconstruction
of the crash, the answer is no. For the benefit of the prosecutor,
defense attorney and mostly the jury, yes it should be included.
These entities provide a visual reference for the lay observer.
When identifying the position of a street lamp for example you
must be familiar with the base point of the symbol to be inserted.
Therein lays the importance of being proficient at mapping. The
rod person and station operator should be equally knowledgeable
with the mapping process and map software to be utilized. During
training this aspect of forensic mapping is stressed to students.
I still shudder when I hear a student say, "I'll just hold
the pole." Well, my friend, you have just volunteered for
the most hazardous, stressful and important job at the incident.
I would rather erect a 5000 piece 3D jigsaw puzzle than volunteer
with the notion of simplicity in my head. I would jump at the
chance however, prepared with proper training, a plan in mind,
and established communication with the set operator.
A
symbol is a group of entities locked together to form a single
entity. This new entity is saved in the forensic library to be
recalled over and over again. Every time you reuse a symbol, you
saved its creation time. The map code library has no idea that
the code FH is representative of a fire hydrant. The software
only knows the FH code holds an instruction to draw 187 entities
locked together. This object however, visually represents what
the human observer will recognize as a fire hydrant. A hydrant
is usually measured to its center, performed by a distance then
offset angle measurement. The measurement is taken to the center
because it is a common position between all hydrants and the base
point for the symbol is the center. A street lamp could have a
base point at the center of the entity. At the physical location
of the lamp pole being measured it is not physically possible
within acceptable time constraints to identify the physical center
of the real world lamp pole.
|
In Figure
17 we examine two pairs of similar symbols - the topographic
view of a lamp pole and fire hydrant. These are simple objects
to visualize and consider where a real world measurement
should be recorded to accurately and automatically insert
a symbol. Center of the vertical pole is a good bet. If
the draftsman who created the lamp pole selected the center
of the object when it was saved, then without editing, the
center of the object must be measured.
Figure
18 displays the same four objects in a 3D view. Measuring
to the center of the hydrant is not a complicated task.
The lamp pole however is more complicated if the base point
is the center of the object and not the center of the vertical
pole. Even more so if we are dealing with a 3D pole with
a base point centered on the horizontal arm.
The
obvious question is how much extraneous mapping is enough.
With experience a properly trained mapping technician will
be competent to make that decision as a part of the mapping
team.
A map
technician should utilize restraint in the use of symbols.
A symbol is a graphic representation of a real world object.
A model is a scaled version of a real work object. In my
review of map projects from North and South America, Western
Europe and the South Pacific, the use of symbols as models
is prolific. So much so that it borders on jeopardizing
the technical accuracy of the project. The method used to
create the symbol is usually not known by the end user.
The symbol could be a simple drawing; it may be measured
with a Total Station, Laser Scanner, or Faroarm, (Figure
19) - each having a different accuracy.
The
use of symbols is not discouraged as long as the difference
is clearly articulated in an accompanying report or diagram
legend. A mapping technician usually struggles with the
"how much" question at an incident scene. At a
crash scene, determine an estimate for the approach speeds
of the vehicles, convert to feet per second, and multiple
by 10. This simple rule will provide you with about 10 seconds
of approach roadway. Keep in mind, a vehicle pulling out
from a stop will need much less approach than does a vehicle
traveling at 90 feet per second. This rule usually provides
sufficient roadway for a time distance analysis. At a crime
scene a good rule is to include all physical evidence and
positions of witnesses. Where was the witness when he heard
three pops, turned around and saw the victim get shot? Every
incident in the investigators mind should be a crime scene
until the investigation is complete or nearly so - at least
until the analysis is far enough along to make the call
- crime or accident.
|
Figure 17 Topographic View
Figure 18 3D view
|
 In
a final drawing such as Figure 20, keep the map simple. The more
objects you place in the map, the more an observer must visually
scan for the object of your testimony. If you are intent on making
your map an exact duplicate of the site, take a picture. You will
save yourself from headaches, upset stomach and a seat at the
boss's desk explaining why it has taken three days to finish a
technical drawing. Figure 20 was mapped using a Sokkia Set 530R.
The prism mode was utilized for all roadway measurements and offsets
to poles. The street lamp heads were measured using the reflectorless
option. Extra caution while aiming and measuring in the reflectorless
mode is necessary to ensure accurate measuring. Real time mapping
makes any reflectorless measuring a breeze. The site work was
completed in fifty-four minutes totaling 124 measurements.
The
map was downloaded from a Panasonic with Evidence Recorder via
Active Sync to MapScenes PRO. The transfer using the download
Wizard takes all of two minutes with a slow computer. With the
2.55 gig hz at my desk, I don't have time to fill my coffee cup.
Forensic
Mapping certainly has challenges associated with it. The process
is truly an art and science. This practice should never be confused
with Surveying, a license-required position. Once your initial
training is complete it is recommended that you participate in
a Field Demonstrated Proficiency Program. Once a week for eight
weeks, you will perform a test map. These locations should be
selected based upon your highest crash locations. The practice
exercises can be printed and used for less serious crashes by
adding scene evidence. It is important to share with coworkers
what you would like them to use as a reference point. At the end
of eight weeks, your confidence level will be increased, your
setup time will be a few minutes, your mapping skills will be
sharp and you will be prepared to meet the challenges of mapping
during peak stress periods. The benefits of applying this technology
to crash and crime scenes will mount as you and your agency gain
experience. Your equipment, skills and application will survive
a judicial challenge and keep the scales of justice balanced.
Link
to Part I and II
President
of McKinzie & Associates, Steve is an active reconstructionist
specializing in commercial vehicle collision reconstruction and
Forensic Mapping.
Return
to March 2003 Newsletter
|
