Highway Asphalt Pavement Degradation Classification using Deep Learning and Computer Vision

Local Chapter Port Harcourt, Nigeria Chapter

Coordinated byNigeria ,

Status: Completed

Project Duration: 08 Mar 2023 - 01 Jul 2023

Open Source resources available from this project

Project background.

Asphalt Pavement Degradation is a common problem on Nigerian highways. Major roads linking streets and even big cities face a serious challenge of bad roads due to the state of the road. Professionals in the field have agreed that the best way to reconstruct a road is to first know the type of degradation which provides data for the best decision to take. If this is ignored, fixing a degraded highway becomes blind and less effective.

The regular way of identifying a degraded pavement will be for Engineers to do onsite surveys. The different types of pavement degradation include Linear Cracks, Crocodile Cracks, Potholes (most prevalent in Nigeria), Fatigue Cracks, Blowouts, Reflection Cracks, sinkholes, Block Cracks, Rutting, and Ravelling. This process stands a chance of being automated.

The problem.

Highway pavement degradation is a prevalent situation in Nigeria. The regular operation will be for Engineers or observation onsite. Due to the large land mass of Nigeria and the harsh weather condition, it becomes stressful and time-consuming to use the regular way of onsite observation. This could also disrupt traffic on the highway.
Deep learning and computer vision could be used to identify and also classify this degraded asphalt pavement using drone images all in real-time.
Though much research has tried to build models, this project will try to increase the classes and also address Highways in Nigeria.

Project goals.

The goals of the project are:In this project, the Port Harcourt Chapter team aims to develop a Deep Learning model that will identify and classify different types of asphalt pavement degradations. The project's primary goal is to accurately identify asphalt pavement degradation.With a duration of 8-weeks, this project aims to:Data Collection and Exploratory Data Analysis Preprocessing/annotation Feature Extraction Model Development and Training Evaluate Model App development/Deployment

Project plan.

  • Week 1


  • Week 2

    Data acquisition

  • Week 3

    Data Understanding/Literature Review

  • Week 4

    Selection of technologies and stacks

  • Week 5

    Data Cleaning/Annotation

  • Week 6

    Data Cleaning/Annotation

  • Week 7

    Model Building

  • Week 8


Learning outcomes.

Deep learning, image processing, computer vision, team work, problem solving

Share project on: