DocumentCode :
3681990
Title :
Stabilization of 3D Pavement Images for Pothole Metrology Using the Kalman Filter
Author :
Abdullah Rasheed;Khurram Kamal;Tayyab Zafar;Senthan Mathavan;Mujib Rahman
Author_Institution :
Nat. Univ. of Sci. &
fYear :
2015
Firstpage :
2671
Lastpage :
2676
Abstract :
Roads are scanned with automated imagers for distresses on a regular basis. Motion of imaging platforms introduces instability in pavement images during image acquisition process, thus, provides less accurate measurements. In this paper, a novel approach is proposed to stabilize 3D pavement images using the Kalman filter with affine transformations as a state space model. The vibration of the imaging platform is taken using a simulated accelerometer and used as a significant feature to measure the effects of instability in pavement images. A Simulink model is presented in this regard to demonstrate the stabilization technique. This paper is an extension of our previous work on metrology and visualization of potholes using Kinect sensor. The effects of vibration for a displacement range of 0.2-1 mm are studied for pure translations in 3D pavement images. The Kalman filter shows a promising future for stabilization of 3D pavement images. Calculations using values from Kalman filter show a reduction in error from 5.35% to 3.76% and from 5.38% to 3.09% for volume and perimeter estimations respectively.
Keywords :
"Kalman filters","Three-dimensional displays","Mathematical model","Robot sensing systems","Estimation","Cameras"
Publisher :
ieee
Conference_Titel :
Intelligent Transportation Systems (ITSC), 2015 IEEE 18th International Conference on
ISSN :
2153-0009
Electronic_ISBN :
2153-0017
Type :
conf
DOI :
10.1109/ITSC.2015.429
Filename :
7313521
Link To Document :
بازگشت