An imaging system that autonomously monitors lighting patterns with application to airport lighting

Abstract

This paper presents a novel measurement system that assesses the uniformity of a complete airport lighting installation. The system improves safety with regard to aircraft landing procedures by ensuring airport lighting is properly maintained and conforms to current standards and recommendations laid down by the International Civil Aviation Organisation.The measuring device consists of a CMOS vision sensor with associated lens system fitted to the interior of an aircraft. The vision system is capable of capturing sequences of airport lighting images during a normal approach to an aerodrome. These images are then post processed to determine the uniformity of the complete pattern.Airport lighting consists of elevated approach and inset runway luminaires. Each luminaire emits an intensity which is dependant on the angular displacement from the luminaire. For example, during a normal approach a given luminaire will emit its maximum intensity down to its minimum intensity as the aircraft approaches and finally passes over the luminaire. As such, it is possible to predict the intensity that each luminaire within the airport lighting pattern emits, at a given time, during a normal approach. Any luminaires emitting the same intensity can then be banded together for the uniformity analysis.Having derived the theoretical groups of similar luminaires within a standard approach, this information was applied to a sequence of airport lighting images that were recorded during an approach to Belfast International Airport.Since we are looking to determine the uniformity of the pattern, only the total pixel grey level representing each luminaire within each banded group needs to be extracted and tracked through the entire image sequence. Any luminaires which fail to meet the requirements (i.e. a threshold value depending on the performance of the other luminaires in that band) are monitored and reported to the assessor for attention.The extraction and tracking algorithms have been optimised for minimal human intervention. Techniques such as component analysis as well as centre of mass algorithms are used to detect and locate the luminaires. A search algorithm is used to obtain the brightness (total grey level) of each luminaire.For the sample test at Belfast International Airport several luminaires were found that do not output sufficient intensity. As a final conclusion however, the Belfast International lighting pattern is legal and conforms to standards as no two consecutive luminaires fail in the pattern.The techniques used in this paper are novel. No known research exists that couples uniformity of airport lighting with photometrics. A solid basis has been established for future work on monitoring the individual characteristics of the luminaires. This includes colour and intensity measurements. Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/23/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx Vision Sensor(s) Mosnting PItfonn E.g. Focal Length USB2.O Power & Conenes Sofficare 4_+LIIII Pesi...

Fast model-based feature matching technique applied to airport lighting

“…The approach lighting was chosen because of the close proximity of the luminaires and the difficulty this poses for the unique identification of the luminaires. In addition, the approach lighting is elevated from the ground and therefore its performance cannot be assessed using the conventional ground-based methods such as MALMS that have evolved over the past few years [4].…”

“…There is the probability that more than one blob corresponds to the same luminaire since the raw blobs extracted include not only the model luminaires but also those that can arise from the background (such as the horizon) and random image noise. Niblock et al [4] propose that other common problems leading to mismatches between the model and image data occur from factors such as inadequate sensor resolution, reflections causing stray noise, luminaires leaving and entering the field of view of the sensor and occlusions, where a luminaire is obstructed by an obstacle from the field of view of the sensor. In any case, the closest blob is always selected as shown in (12) and the missing blobs are thus identified.…”

“…Success rate, shown by (4), is used to measure the quality of the matching results. This compares the number of successfully identified features (S E ) to the number of expected features (S T ), that is, the number of luminaires that should be present in the image…”

“…All future operations are performed on image data that have been transformed relative to the assigned orientation, scale and location of each feature, thereby providing invariance to these transformations. 4. Keypoint descriptor.…”

“…[4]. Over the years, several land-based measurement systems such as the mobile airfield light monitoring system [MALMS capable of assessing the performance of inset runway lighting using light meters and cameras, respectively.…”