Autonomous Ground Systems CROSS CORRELATING GROUND-LEVEL PANORAMAS - - PowerPoint PPT Presentation

▶
autonomous ground systems
SMART_READER_LITE
LIVE PREVIEW

Autonomous Ground Systems CROSS CORRELATING GROUND-LEVEL PANORAMAS - - PowerPoint PPT Presentation

Jul 26, 2023 245 likes •414 views

Autonomous Ground Systems CROSS CORRELATING GROUND-LEVEL PANORAMAS WITH SATELLITE IMAGERY FOR GPS-DENIED LOCALIZATION OF AUTONOMOUS GROUND VEHICLES Calvin Cheung Stanley Baek U.S. Army Tank Department of Electrical and Computer Army

slide-1
SLIDE 1

Autonomous Ground Systems

8/9/2018

CROSS CORRELATING GROUND-LEVEL PANORAMAS WITH SATELLITE IMAGERY FOR GPS-DENIED LOCALIZATION OF AUTONOMOUS GROUND VEHICLES

DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited.

Calvin Cheung U.S. Army Tank Army Research Development Engineering Center Warren, MI Stanley Baek Department of Electrical and Computer Engineering University of Michigan Dearborn, MI

slide-2
SLIDE 2

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 2

  • Localization – Where am I?

INTRODUCTION

???

slide-3
SLIDE 3

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 3

  • Localization - GPS

INTRODUCTION

slide-4
SLIDE 4

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 4

  • Localization – GPS Degradation

INTRODUCTION

  • Occlusion
  • Spoofing
  • Denial
slide-5
SLIDE 5

Autonomous Ground Systems

8/9/2018 5

  • Localization – GPS-Denied Localization

INTRODUCTION

slide-6
SLIDE 6

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 6

RELATED WORK

Simultaneous Localization and Mapping

*Image from http://www.pirobot.org/blog/0015/

Different Sensing Modalities

*Cornick, Koechling, Byron, Zheng. “Localizing Ground Penetrating RADAR: A Step Toward Robust Autonomous Ground Vehicle Localization”

Landmark Recognition SIFT Matching

slide-7
SLIDE 7

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 7

LOCALIZATION PROCESS

  • 1. Capture a spherical panoramic image on a vehicle and warp it to produce a top-down

view.

slide-8
SLIDE 8

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 8

LOCALIZATION PROCESS

  • 2. Run Canny edge detection on both the top-down view and the satellite image and

perform normalized cross correlation on the resulting images.

slide-9
SLIDE 9

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 9

LOCALIZATION PROCESS

  • 3. Filter matches based on distance from the last known good point.
  • 4. Apply Kalman filter step to matched point and report the location.
slide-10
SLIDE 10

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 10

  • Panoramic images downloaded from Google Street view
  • Suburb of Sonoma Peak chosen to test localization process

EXPERIMENT

slide-11
SLIDE 11

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 11

RESULTS

: Ground Truth : Cross-correlation Matches : Kalman Filtered Results.

slide-12
SLIDE 12

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 12

RESULTS

slide-13
SLIDE 13

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 13

RESULTS

RMSE for CC along East/ West (m) RMSE for CC along North/South (m) RMSE for KF along East/ West (m) RMSE for KF along North/South (m) KF Mean Distance from Truth (m) Distance Filtering

72.32 480.33 71.97 398.49 253.78

Canny Edge Detection

97.72 194.70 36.33 136.21 115.65

Canny Edge Detection, Distance Filtering

36.47 81.53 20.69 52.91 46.95

Localization Performance

slide-14
SLIDE 14

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 14

  • Developed a method of GPS-denied localization by performing normalized

cross-correlation on satellite imagery and warped spherical panoramic

  • Using map data from Google Maps and Google Street View, we

demonstrated localization ~50 m of ground truth

  • Method can work with existing GPS solutions (<7.8 m accuracy) to help

maintain coordinates during intermittent GPS drops

  • In complete GPS-denied environments, the localization process can

provide a great benefit for AGVs looking to approximate position

CONCLUSIONS

slide-15
SLIDE 15

Autonomous Ground Systems

8/9/2018 DISTRIBUTION STATEMENT A. Approved for public release; distribution unlimited. 15

  • Questions/comments?

THANKS!