Road Pavement Condition Monitoring by Embedded Crowdsensing
Android crowdsensing app classifying eight types of road degradation using camera and accelerometer data.
May 2020 – August 2020 · ATLSS Engineering Research Center, Lehigh REU
I helped develop an Android mobile application that crowdsources road-pavement data using camera and accelerometer signals. The app uses a MobileNet-SSD image classifier to identify eight types of road degradation at various coordinates with a confidence level of 60%.
I contributed to a technical report that was published, and presented the project at the CIAMTIS conference.
