Spatial Resolution Ensured Configuration Planning for Visual Inspection Using Pan-Tilt-Zoom Camera-Equipped Robots

Abstract

We propose a novel pan-tilt-zoom (PTZ) camera configuration planning method to improve data collection for structure visual inspections. This method plans a set of configurations for PTZ camera-equipped robots to collect visual data of target surfaces at the required spatial resolution. Unlike previous methods that focus solely on optimizing camera positions to cover different areas of the target, our method fully leverages all the degrees of freedom of PTZ cameras. Our method allows the robot to completely cover targets with far less movement by adjusting the camera direction and zoom level. Experiments on bridge inspections using a PTZ ground inspection platform have validated that our method ensured both the completeness and spatial resolution of the collected data. Furthermore, drone-based inspection simulations have demonstrated that our method avoids unnecessary robot movement and reduces inspection time by up to 38% compared to the previous algorithm, significantly enhancing inspection efficiency.

Keywords

• Motion and path planning
• aerial systems: perception and autonomy
• field robots
• robotics and automation in construction
• autonomous agents
• reactive and sensor-based planning