Road Pit hole detection using OpenCV - Part 1

Introduction

This article is about detecting pit holes on road.It analyses images identifies craks & distresses on road

Background

My friend is a highway engineer. Worked in Chennai,India.He did lot of manual work on identifing pit holes.Survey is done manually employing engineers Measure distresses based on length, depth and area

Java code sniplet

  FeatureDetector blobDetector = FeatureDetector.

create(FeatureDetector.ORB);
  blobDetector.detect(srcImgMat, matOfKeyPoints);

  System.out.println("Detected " +

 matOfKeyPoints.size() + " blobs in the image");

After that i have identified the keypoints, i have plotted them in the original image We take a video of the road, we will apply same logic to all the frame. This will be done using VideoCapture api of opencv. Then we calculate the SQFT of the distress/crack. Then we calculate the cost of filling the distress. This would reduce rework of laying road and it will last longer. We can make use of tensorflow to identify the cracks. I will post an article in part 2 article.
Hope you liked it