Val Noronha University of California, Santa Barbara Integrating RS into Transportation Information Systems.

Val NoronhaUniversity of California, Santa Barbara

Integrating RS into Transportation

Information Systems

STELLA/STAR 2002-01-17 #2 N C R S T

Why this engineering stuff?

Methodological research• DOT as customer: GI-science brings

operational efficiencies• Academics (social scientists) as end-

users: facilitates DOT data, e.g. access, safety

Hands across the sea: idea exchange


Background

U.S. DOT — NASAPartnership

Transportation Legislation “TEA-21”

Environment

Flows Hazards

Infrastructure

4 universities

4 TAPs


NCRST Philosophy

Rapid evolution of research into commercial products, practice

User consultation and outreach Partnerships: industry,

international


Research Production Line

ProblemSensor

Image processing

Data integration

Decision/solution

Data management

Needs analysisSensor technology

Image processing tech

Analysis/process model

Education & outreach

Data models,communication


NCRST—Infrastructure

Asset management (esp. location referencing)

ITS Corridor/facility location Security

C E N T E R L I N E

M A P P I N G



Centerline databases — families

Pro

duce

rsU

sers

ITS

State, localDOT

USGS1:24K

State, local

Engineering

DesignCAD

Mapquest

TIGER

GDTNavtech

CVO/Logistics

RS RS RS


Early Thinking

RS is not a monolithic solution; many types of imagery many approaches

No centerline solution is the “right” answer for all needs; each addresses a niche• scale• rural vs urban

Initial plan: validate RS with GPS


Imagery Resources

AVIRIS

Soft copyphotogrammetry

Landsat TM

Spatial resolution

Spect

ral re

solu

tion

IKONOS

Material classification

Edge detection


Lessons Learnt

There are many approaches No RS centerline solution is the

answer to all our needs; each addresses a niche

Manual process with some automation, or largely automated process with some user intervention


The GPS Challenge


GPS Results — lane resolution


GPS for Hwy Ops

Hi end

Lo end compatibility


Centerlines from RS — conclusion

RS not a fully automated solution … yet Augments other methods (e.g. GPS)

effectively, esp. visual verification Good for small areas: 1:5K soft-copy

photogrammetry Good for large areas, 1:50K to 1:100K …

global databases, sustainable policy, etc Research is promising, worth pursuing


Centerline databases — families

Pro

duce

rsU

sers

ITS

State, localDOT

USGS1:24K

State, local

Engineering

DesignCAD

Mapquest

TIGER

GDTNavtech

CVO/Logistics

RS RS RS


Data exchange support

UNETRANS — with ESRI NSDI Road Identification Standard NCHRP 20-27(3) — led by UW Inter-state partnerships


UNETRANS

A I R P O R T

P L A N N I N G


Airport Planning

FAA priorities• 3-D Airport Layout Planning• 3-D Airport Approach Planning

LIDAR projects• St Petersburg, FL• Plant City, FL


3-D Airport Approach Planning



New Projects

Security of infrastructure Corridor alignment Pavement condition (hyperspectral) LIDAR and photogrammetry


OrbImage — Imagery Access

Virginia — widening Rte 1 from Woodrow Wilson bridge to Dale City


Tetra Tech — multi-modal transfers

Land use classification — methodology test


Counting containers — methodology development




Washington Blvd corridor ...


… objects in 15 cm imagery



Boston — Big Dig impacts


www.ncgia.org• CSISS• VITAL• NCRST

Keep in touch!

Val Noronha University of California, Santa Barbara Integrating RS into Transportation Information Systems.

Documents

n c r s t stellastar

gps slide

c e n t e r

communication slide

international slide

gps challenge slide

p p i n g slide

user intervention slide