Download presentation
Presentation is loading. Please wait.
Published byDerrick Grant Modified over 9 years ago
1
Val Noronha University of California, Santa Barbara Transportation Centerlines: Case Studies from California and Iowa
2
N C R S T RSPA/FDOT 2004-12-02 #2 © University of California. All rights reserved. Structure U.S. DOT — NASA Partnership Transportation Legislation “TEA-21” EnvironmentFlowsHazardsInfrastructure 4 universities 4 TAPs
3
N C R S T RSPA/FDOT 2004-12-02 #3 © University of California. All rights reserved. Outline What is a centerline? Applications, requirements, status Survey Methods Remote sensing: eyeball, analytical GPS: high-end, low-end Photogrammetry Tomorrow’s precision standards
4
N C R S T RSPA/FDOT 2004-12-02 #4 © University of California. All rights reserved.
5
N C R S T RSPA/FDOT 2004-12-02 #5 © University of California. All rights reserved.
6
N C R S T RSPA/FDOT 2004-12-02 #6 © University of California. All rights reserved. The Grapevine Northbound and southbound carriageways of I-5 switch places for 6 km along the San Andreas Fault
7
N C R S T RSPA/FDOT 2004-12-02 #7 © University of California. All rights reserved.
8
N C R S T RSPA/FDOT 2004-12-02 #8 © University of California. All rights reserved. Where is the Gore point?
9
N C R S T RSPA/FDOT 2004-12-02 #9 © University of California. All rights reserved. Centerline Applications Who’s paying?
12
N C R S T RSPA/FDOT 2004-12-02 #12 © University of California. All rights reserved. Centerline databases — families Producers Users ITS State, local DOT USGS 1:24K State, local Engineering Design CAD Mapquest TIGER GDT Navtech CVO/Logistics RS
13
N C R S T RSPA/FDOT 2004-12-02 #13 © University of California. All rights reserved. Facets of centerline data Existence: omission/commission Geometric accuracy (2-D) Linear accuracy Attributes name and address range, speed limit lane width, shoulder width … Topology Data model
14
N C R S T RSPA/FDOT 2004-12-02 #14 © University of California. All rights reserved.
15
N C R S T RSPA/FDOT 2004-12-02 #15 © University of California. All rights reserved.
16
N C R S T RSPA/FDOT 2004-12-02 #16 © University of California. All rights reserved.
17
N C R S T RSPA/FDOT 2004-12-02 #17 © University of California. All rights reserved.
18
N C R S T RSPA/FDOT 2004-12-02 #18 © University of California. All rights reserved.
19
N C R S T RSPA/FDOT 2004-12-02 #19 © University of California. All rights reserved.
20
N C R S T RSPA/FDOT 2004-12-02 #20 © University of California. All rights reserved.
21
N C R S T RSPA/FDOT 2004-12-02 #21 © University of California. All rights reserved.
22
N C R S T RSPA/FDOT 2004-12-02 #22 © University of California. All rights reserved.
23
N C R S T RSPA/FDOT 2004-12-02 #23 © University of California. All rights reserved.
24
N C R S T RSPA/FDOT 2004-12-02 #24 © University of California. All rights reserved.
25
N C R S T RSPA/FDOT 2004-12-02 #25 © University of California. All rights reserved.
26
N C R S T RSPA/FDOT 2004-12-02 #26 © University of California. All rights reserved.
27
N C R S T RSPA/FDOT 2004-12-02 #27 © University of California. All rights reserved. Name Discrepancies
28
N C R S T RSPA/FDOT 2004-12-02 #28 © University of California. All rights reserved. Solutions—Geometry Remote sensing GPS high-end dedicated multiple vehicles? Photogrammetry Mosaic engineering drawings
29
R E M O T E S E N S I N G
31
N C R S T RSPA/FDOT 2004-12-02 #31 © University of California. All rights reserved. Hyperspectral process 1 MESMA (Dar) Q-treeVectorize Additional steps: clean, conflate
32
N C R S T RSPA/FDOT 2004-12-02 #32 © University of California. All rights reserved. MESMA (Dar) Thinning, Filters Comparison Additional steps: clean, conflate Hyperspectral process 2
33
T H E G P S C H A L L E N G E
34
N C R S T RSPA/FDOT 2004-12-02 #34 © University of California. All rights reserved. High end (“SOP”) GPS ARAN Output: GPS every 10 m — not topological centerlines Cost: x 10 4 (service) Roadware Inc
35
N C R S T RSPA/FDOT 2004-12-02 #35 © University of California. All rights reserved.
36
N C R S T RSPA/FDOT 2004-12-02 #36 © University of California. All rights reserved. ARAN output
37
N C R S T RSPA/FDOT 2004-12-02 #37 © University of California. All rights reserved. Problems w/ Dedicated GPS Need to get it right the first time Verification? IMU helps ($$$)
38
N C R S T RSPA/FDOT 2004-12-02 #38 © University of California. All rights reserved. The Consumer-GPS Challenge
39
N C R S T RSPA/FDOT 2004-12-02 #39 © University of California. All rights reserved. GPS Results — lane resolution
40
N C R S T RSPA/FDOT 2004-12-02 #40 © University of California. All rights reserved. The one to beat … $150 at Staples Convenience Price
41
T H E G R E A T G P S - L I N E A R COMPATIBILITY C H A L L E N G E ?
42
N C R S T RSPA/FDOT 2004-12-02 #42 © University of California. All rights reserved. LX for professionals DMI: accurate to 1 m Linear referencing: “US101 — 20.25”
43
N C R S T RSPA/FDOT 2004-12-02 #43 © University of California. All rights reserved. A 7075 C 7472 E 7330 DMI 7477 GPS 7446
44
N C R S T RSPA/FDOT 2004-12-02 #44 © University of California. All rights reserved. A Myth about Linear vs 2D Distance over the hill is greater: need to correct for elevation Difference = 0.25% (on a 7% grade)
45
N C R S T RSPA/FDOT 2004-12-02 #45 © University of California. All rights reserved. What Really Happens on the Hill Inaccurate (x,y) geometry shortchanges length up to 20%
46
N C R S T RSPA/FDOT 2004-12-02 #46 © University of California. All rights reserved. Linear Accuracy Tests — Setting
47
N C R S T RSPA/FDOT 2004-12-02 #47 © University of California. All rights reserved. Linear Accuracy Tests — Setting ~10 km road, rises 800m, average 8% grade Numerous hairpins Some dense tree canopy and partial GPS occlusion Remote sensing would have difficulty with this
48
N C R S T RSPA/FDOT 2004-12-02 #48 © University of California. All rights reserved. Linear Accuracy Test — Results $150-GPS length close to DMI uphill: 1 ~ 1.5% downhill: 0.5% Uphill/downhill difference Simulated: opposing lane geometry: 0.02-0.05% GPS: 0.2% (higher downhill speed?) DMI: 1.5% (affected by engine rev?) Conclusion (provisional, more study needed) GPS just as accurate (±0.5%), more consistent
49
N C R S T RSPA/FDOT 2004-12-02 #49 © University of California. All rights reserved.
50
N C R S T RSPA/FDOT 2004-12-02 #50 © University of California. All rights reserved. Will GPS replace LR?
51
N C R S T RSPA/FDOT 2004-12-02 #51 © University of California. All rights reserved. Big Picture
52
N C R S T RSPA/FDOT 2004-12-02 #52 © University of California. All rights reserved. Manual Softcopy
53
N C R S T RSPA/FDOT 2004-12-02 #53 © University of California. All rights reserved. UNETRANS
55
N C R S T RSPA/FDOT 2004-12-02 #55 © University of California. All rights reserved. The Future ITS lane departure warning autonomous control and platooning Sub-decimeter accuracy Millions of sensors
56
N C R S T RSPA/FDOT 2004-12-02 #56 © University of California. All rights reserved. Tomorrow’s Centerlines Car navigation system automatically updates local map while you gas up
57
N C R S T RSPA/FDOT 2004-12-02 #57 © University of California. All rights reserved. Conclusions No single solution to all requirements Showing promise Remote sensing GPS Photogrammetry Attribution, conflation, modeling … and data exchange
59
1 www.ncgia.org
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.