Passive obstacle detection system (PODS) for wire detection
Author(s) -
John N. Sanders-Reed,
Dennis J. Yelton,
Christian Witt,
Ralph R. Galetti
Publication year - 2009
Publication title -
proceedings of spie, the international society for optical engineering/proceedings of spie
Language(s) - English
Resource type - Conference proceedings
SCImago Journal Rank - 0.192
H-Index - 176
eISSN - 1996-756X
pISSN - 0277-786X
DOI - 10.1117/12.815543
Subject(s) - pixel , obstacle , extremely high frequency , image processing , remote sensing , computer science , computer vision , artificial intelligence , object detection , optics , image (mathematics) , telecommunications , physics , geology , geography , pattern recognition (psychology) , archaeology
Boeing has developed algorithms and processing to detect power lines and cables in passive imagery from a wide variety of different sources. The algorithm has been demonstrated with imagery from visible, medium and long wave infra-red (MWIR and LWIR), and Passive MilliMeter Wave (PMMW) sensors. Flight demonstrations of the real-time system have been performed with both visible and LWIR image sensors. In the flight demonstrations, an LWIR image sensor was used for both day and night detection in both rural and urban settings with detection ranges in excess of 15,000'. The processing system is capable of processing image sizes up to 1024x1024 pixels at 30 frames per second. A key component of aircraft obstacle avoidance is the detection of power lines and cables. Within the general area of obstacle detection, power lines and cables present some unique detection challenges. Due to their small diameter (typically less than 1.25") they do not show up in digital terrain maps and can be very difficult to detect visually, especially when seen against complex natural backgrounds. While radio towers, smoke stacks, cranes, and other tall, narrow man-made objects typically have warning lights to improve their visibility, most power lines and their supporting pylons and poles have no such visibility enhancements. In order to support a wide range of operating conditions, it is necessary to be able to detect power lines and cables during both day and night, and in reduced visibility conditions. Required detection range is a function of aircraft speed and maneuverability. At 150 kt (typical top speed for a helicopter) the aircraft covers 250 ft/sec while at a fixed wing speed of 250 kt an aircraft covers 420 ft/sec. If one uses a minimum reaction time of 3 seconds, then for a helicopter flying nap of the earth at full speed, the minimum useful detection range is 750 ft while for a fixed wing aircraft it may be more like 1300 ft. A much more comfortable time lag of 10 sec gives desired detection ranges of 2500 ft and 4200 ft respectively. High voltage power lines tend to be approximately 1" or 1.25" in diameter while local distribution lines are more typically ½" in diameter. However, the smaller distribution lines tend to be much closer to the ground and therefore pose much less of a flight hazard. A number of different approaches are under development. These include sensors to detect electro-magnetic emissions from active power lines and various forms of lidar and radar. Boeing has been developing real-time passive wire detection systems since 1998, resulting in 1 patent and 2 published patent applications (1-3). Passive wire detection has several distinct advantages over other approaches but also faces significant technical challenges. Unlike detectors of electro-magnetic emissions, both lidar/radar and passive detection techniques can detect both powered and un-powered lines and can provide precision location of the wires and cables. In comparison with active systems, the presence and operation of a passive system is completely undetectable. While lidar systems may be claimed to be Low Probability of Intercept / Low Probability of Detection (LPI/LPD), passive systems are by definition Zero Probability of Intercept / Zero Probability of Detection (ZPI/ZPD). Further, active systems are limited in range by their transmitter power and the cross section of the object to be detected. Initial detection range, with zero false alarms, for the PODS wire detection system is 15,000 ft (36 sec at 250 kt). Detection strength increases to detection in over 50% of the image frames by 11,000 ft (26 sec at 250 kt) and continuous detection (100% of frames) by 2300 ft (5 sec at 250 kt).
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom