Presentation is loading. Please wait.

Presentation is loading. Please wait.

ASSESSING PAVED ROAD SURFACE CONDITION WITH HIGH-RESOLUTION SATELLITE IMAGERY William J. Emery (University of Colorado) Ashwin Yerasi (University of Colorado)

Similar presentations


Presentation on theme: "ASSESSING PAVED ROAD SURFACE CONDITION WITH HIGH-RESOLUTION SATELLITE IMAGERY William J. Emery (University of Colorado) Ashwin Yerasi (University of Colorado)"— Presentation transcript:

1 ASSESSING PAVED ROAD SURFACE CONDITION WITH HIGH-RESOLUTION SATELLITE IMAGERY William J. Emery (University of Colorado) Ashwin Yerasi (University of Colorado) Nathan Longbotham (DigitalGlobe) Fabio Pacifici (DigitalGlobe) 1IGARSS 2014

2 Outline Background Road Quality Assessment Road Asphalt Identification Final Remarks 2IGARSS 2014

3 Background 3IGARSS 2014

4 Motivation In situ surveillance of paved road surfaces –Primarily performed manually –Slow and tedious –Limited coverage Remote sensing of paved road surfaces –Primarily performed automatically –Comparatively efficient –Large area coverage Latter technique can be used as a precursor or compliment to the former 4IGARSS 2014

5 In Situ Data Standard road surface parameters of interest –Roughness (IRI) –Rutting –Cracking (fatigue, etc.) Interpretation of measurements varies by planning organization In general, road condition is rated holistically –Good, fair, poor –High, moderate, low drivability –Etc. Road Quality Surveillance Van (Pathway Services Inc.) 5IGARSS 2014

6 Remotely Sensed Data Provided by DigitalGlobe Collected by WorldView-2 spacecraft Panchromatic imagery –1 band ( nm) –Spatial resolution ~0.5 m –11-bit digital numbers Multispectral imagery –8 bands ( nm) –Spatial resolution ~2 m –11-bit digital numbers WorldView-2 (DigitalGlobe) 6IGARSS 2014

7 Road Quality Assessment 7IGARSS 2014

8 Asphalt Degradation Lighter and less uniform appearance correlated with degradation Road quality thus potentially assessable through texture analysis of imagery Asphalt Spectra (M. Herold) 8 Fair Road (Boulder County)Poor Road (Boulder County) Good Road (Boulder County) IGARSS 2014

9 Road Quality Assessment Overview 9 Panchromatic Imagery Road Asphalt ROIs Texture Filtered Imagery (3 Features) Asphalt Pixel Statistics Road Quality IGARSS 2014

10 Digital Number 21B 115A24A 10 Colorado Springs Highways IGARSS 2014

11 Digital Number GoodFairPoor Mean STD IGARSS 2014

12 Data Range 21B 115A24A 12 Colorado Springs Highways IGARSS 2014

13 Data Range GoodFairPoor Mean STD IGARSS 2014

14 Variance 21B 115A24A 14 Colorado Springs Highways IGARSS 2014

15 Variance GoodFairPoor Mean STD IGARSS 2014

16 Entropy 21B 115A24A 16 Colorado Springs Highways IGARSS 2014

17 Entropy GoodFairPoor Mean STD IGARSS 2014

18 WorldView-2 Sensor Noise Analysis GoodFairPoor Data DN Mean Data DN STD Sensor Noise IGARSS 2014 WV-2 Sensor Noise (DigitalGlobe) Colorado Springs Highways

19 Road Asphalt Identification 19IGARSS 2014

20 Road Asphalt Identification Overview 20 Panchromatic Imagery Multispectral Imagery Pansharpened Imagery (8 Bands) Road Asphalt ROIs Texture Filtered Imagery (3 Features) IGARSS 2014 OpenStreetMap Shapefiles

21 Road Identification Must disregard non-road features in scenery Use OpenStreetMap shapefiles as mask Original Scene Masked Scene Original Scene with OSM Shapefile 21IGARSS 2014

22 Asphalt Identification Must distinguish asphalt from non-asphalt features in roads – Vehicles, paint, shadows, etc. 11 total dimensions contained in image pixels –8 spectral, 3 texture Training set manually selected –Asphalt vs. non-asphalt Random forest classification –Cohen’s kappa coefficient of ~0.91 obtained from experimental trials Asphalt Vehicle Paint Shadow 22IGARSS 2014

23 Final Remarks 23IGARSS 2014

24 Conclusions Road asphalt can be identified from high-resolution satellite imagery For the data analyzed, road asphalt becomes lighter in panchromatic grayscale shade as it degrades –Digital number increases For the data analyzed, road asphalt becomes less uniform in texture as it degrades –Data range increases –Variance increases –Entropy increases These apparent qualities can potentially be used to assess road pavement condition via satellite remote sensing 24 IGARSS 2014

25 Questions IGARSS

26 Backup Slides 26IGARSS 2014

27 Pansharpening Panchromatic Multispectral Pansharpened 27IGARSS 2014

28 Occurrence-Based Texture Filtering I(1,1)I(1,2)I(1,3)I(1,4)I(1,5) I(2,1)I(2,2)I(2,3)I(2,4)I(2,5) I(3,1)I(3,2)I(3,3)I(3,4)I(3,5) I(4,1)I(4,2)I(4,3)I(4,4)I(4,5) I(5,1)I(5,2)I(5,3)I(5,4)I(5,5) T(2,2) T(2,3) T(2,4) T(3,2) T(3,3)T(3,4) T(4,2) T(4,3) T(4,4) Original Image Filtered Image 28IGARSS 2014

29 Occurrence-Based Texture Filtering Digital Number (Original Data) Data Range Variance Entropy 29IGARSS 2014

30 Digital Number 30 Loveland Highways 25A 392A 287C IGARSS 2014

31 Digital Number GoodFairPoor Mean STD IGARSS 2014

32 Data Range 32 Loveland Highways 25A 392A 287C IGARSS 2014

33 Data Range GoodFairPoor Mean STD IGARSS 2014

34 Variance 34 Loveland Highways 25A 392A 287C IGARSS 2014

35 Variance GoodFairPoor Mean STD IGARSS 2014

36 Entropy 25A 392A 287C 36 Loveland Highways IGARSS 2014

37 Entropy GoodFairPoor Mean STD0.2 37IGARSS 2014

38 WorldView-2 Sensor Noise Analysis GoodFairPoor Data DN Mean Data DN STD Sensor Noise IGARSS 2014 WV-2 Sensor Noise (DigitalGlobe) Loveland Highways


Download ppt "ASSESSING PAVED ROAD SURFACE CONDITION WITH HIGH-RESOLUTION SATELLITE IMAGERY William J. Emery (University of Colorado) Ashwin Yerasi (University of Colorado)"

Similar presentations


Ads by Google