PLineD: Vision-based power lines detection for Unmanned Aerial Vehicles

Energy-Efficient Industrial Internet of UAVs for Power Line Inspection in Smart Grid

“…The authors in [14] studied a method for transmission line inspection and tracking by employing Hough transform and Kalman filter. In [15], a power line inspection algorithm called PLineD was proposed to address the real-time perception problem. A trajectory scheduling approach for UAVs in a complex 3D environment was designed in [16] by combining GA and particle swarm optimization to minimize the execution time.…”

Real- Time LiDAR-based Power Lines Detection for Unmanned Aerial Vehicles

“…The condition of the power components needs to be checked for both mechanical and electrical faults [12,19], which occur on conductors, insulators (figure 1.2(b)), electric pylons and other power line equipment [20,21]. Damaged or corroded parts and contaminated conductors are some of the problems that can be visually detected [19,20,22], but corona or gap discharges can also indicate problems on the transmission line [20]. The inspections of the power lines traditionally rely on human labour [10][11][12][13], as they are primarily performed on foot by a team driving between spans (space between two consecutive electric pylons) or from a helicopter flying alongside the line.…”

“…The power structures can be installed on complex and harsh environments, which makes the onfoot patrols time-consuming and poorly efficient. Helicopter-based inspections are more time-efficient but are both expensive and taxing on the pilot and operator [13,22,25].…”

“…Synthetic Aperture Radar (SAR) images [26,27], optical satellite images [13,28] or Airborne Laser Scanning (ALS) data [29][30][31] are some of the examples, however, due to the lack of resolution and optical or thermal images, these methods can only be used for mapping pylons and conductors or vegetation monitoring [11]. A proper inspection of the power line elements can be done by using laser scanning data, optical [32] and thermal [33,34] images obtained by UAVs [10,22,35,36], Rolling on Wires Robots (RWRs) [10,37] or even land-based mobile systems [38]. Regarding the diversity of the data, fixed-wing UAVs can only provide information about the upper part of the line, contrasting with land-based systems, that only give information about the lower one.…”

“…Despite the advantages, using a multirotor UAV in a complex and confined environment, as one where a power line can be inserted in, can be challenging for the pilot and the autonomous mission planning. A noisy background (e.g., with vegetation) can make the conductors visually undetectable by the pilot [22]. This arises the need for an autonomous module of obstacle detection and collision avoidance that is capable of actuating in both autonomous and assisted modes [40].…”