文章基本信息

标题：Real Time Image Classification and Detection of Potholes using Adam’s Optimizer with CNN Framework
本地全文：下载
作者：R. Sathya ; Muppala Nikhil ; K.S.S. Shanmukha Srinath 等
期刊名称：International Journal of Early Childhood Special Education
电子版ISSN：1308-5581
出版年度：2022
卷号：14
期号：2
页码：3714-3721
DOI：10.9756/INT-JECSE/V14I2.400
语种：English
出版社：International Journal of Early Childhood Special Education
摘要：It is important to detect potholes during road maintenance. Road surface models are usually created from 2D road images or from 3D computer vision approaches. They are always employed independently of one another. Further, the accuracy of pothole detection remains far from satisfactory. We developed an accurate and efficient pothole detection algorithm using Adam’s optimizer. The first step is identifying damaged and undamaged areas of roads to transform the dense disparity map. Estimating the transformation parameters utilizing dynamic programming and golden section search will achieve greater disparity transformation efficiency. The modified disparity map is then used to locate likely undamaged road portions, by fitting a four-exterior plane with least suitable model. In order to develop the modelling, a normalization of the plane is also necessary added into the plane modelling procedure. According to the experimental results, the proposed system has an accuracy of 98.9% for successful detection.
关键词：Pothole detection;3D modelling;Neural Networks;CNN;Computer vision;Potholes;2D image analysis