Optical flow-based motion-blurred image restoration for transmission line icing airborne monitoring

Recently, image recognition-based unmanned aerial vehicle (UAV) inspection becomes a new trend for electric power transmission line inspection, for what a key issue is to acquire clearer image as much as possible. Thus this paper focus on motion-blurred image restoration of transmission line icing airborne monitoring. For the point spread function estimation, motion blur parameters can't be calculated precisely by common motion-blurred image restoration methods, thus an optical flow-based fusion method is put forward. With the motion information from the optical-flow algorithm, a more accurate point spread function is obtained. Then the improved effect of motion-blurred image restoration is verified experimentally.

I. Introduction

In recent years, power demand is rapidly increasing with the continuous development of economy, and the scale of power grid is also continuously expanded in various countries. Thus the problem of grid security has attracted more attention, for what a key issue is to ensure the safety and reliability of power grid by transmission line inspection. Traditional manual inspection is time-consuming, ineffective and high-risk, which is unable to meet the needs of ‘Smart Grid’. With the rapid development of UAV (Unmanned Air Vehicles), UAV-based inspection has become the new trend for transmission line inspection, considering the agility, high efficiency and speed, etc.