@misc{69,
  keywords = {Pothole, Artificial Intelligence, Machine Learning},
  author = {Peter Daka and Lloyd Hangoma and David Tembo and Harris Shikapande and Lighton Phiri},
  title = {Pothole Detection and Classification System for Zambian Roads: A Case Study of Lusaka},
  abstract = {Potholes are a significant challenge for Zambia's road infrastructure. Posing a serious threat to vehicle safety and effectiveness of road maintenance. In order to enhance road monitoring, this project aimed to develop a real-time pothole detection and classification system utilising web technologies and artificial intelligence (AI). With the use of convolutional neural networks (CNN) for classification and for detection, the system analyses video footage to locate and map potholes using GPS coordinates. In Lusaka, a large amount of data was gathered for the study using both custom datasets and public sources. The technology offers a web-based interactive platform for dynamic road condition monitoring and delivers an 80% detection accuracy. This invention provides a scalable, automated method to improve road safety and maintenance in Zambia by solving the shortcomings of the country’s present human road monitoring procedures.},
  year = {2024},
  pages = {51},
  month = {12/2024},
  publisher = {University of Zambia},
  address = {Lusaka, Zambia},
}