Fig. 9
From: Advanced lightweight deep learning vision framework for efficient pavement damage identification
Damage instance analysis under complex background: the rectangular box represents a correctly detected instance, the elliptical box indicates a missed detection.
