A DESIGN OF ROAD ASPHALT DAMAGE DETECTION SYSTEM THROUGH VIDEO USING FAST FOURIER TRANSFORM

Abstract

The road pavement distress is the main factor that determines the feasibilityroad
conditions. The Assesment of road condition is essential in the management of road
maintenance. Manually, survey to check road conditions conducted by trained surveyor
to do observe directly on the roads. The road which  have traffic  jam,  it could  endanger
the safety of surveyor. The pavement inspection method using the video can be an
alternative way to make it faster, cost effective, and more secure in the implementation of
observation and evaluation of the road conditions. This study used road video record and
then extracted into the image frames. By utilizing several areas the sum value of Fast
Fourier Transform (FFT) is used as features  to  classify  asphalt  road  image  into
categories  based  on  good,  moderate,  minor damage, and severely damage. The results
of research, showed that image classification of pavement  image  with 25  images  as
good, 25  images  as  moderate, 25  images  as  minor damage, and 25 images as
severely damage have accuracy  percentage 98%.