Occlusion Handling in Radar for Detection of Wires and Pylons

A simple method to improve detection and false alarm rates is by tracking objects over successive frames. However, the performance of the tracker is heavily dependent on good continuity of the tracking process. Occlusions of objects are a key challenge for continuous tracking and may cause a tracker to disassociate from its target. In the framework of radar for wires and pylons detection, occlusions of wire and pylons frequently occur, thus deteriorating the effectiveness of a tracker. In this work, the phenomenology of these occlusions is discussed, and a method for occlusion handling is presented. Measurements taken in flight tests with a pulse-Doppler polarimetric radar for obstacle detection demonstrate the effectiveness of this method.

[1]  Rita Cucchiara,et al.  Probabilistic people tracking with appearance models and occlusion classification: The AD-HOC system , 2011, Pattern Recognit. Lett..

[2]  Xin Li,et al.  Contour-based object tracking with occlusion handling in video acquired using mobile cameras , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Dima Damen,et al.  Recognizing linked events: Searching the space of feasible explanations , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Thia Kirubarajan,et al.  Performance measures for multiple target tracking problems , 2011, 14th International Conference on Information Fusion.

[5]  E. Collett.  Field Guide to Polarization , 2005 .

[7]  Reiner S. Thomä,et al.  Dynamic-occlusion likelihood incorporation in a PHD filter based range-only tracking system , 2015, 2015 18th International Conference on Information Fusion (Fusion).

[8]  Sonia Vatta,et al.  Occlusion Detection and Handling: A Review , 2015 .

[9]  Ralph Helmar Rasshofer,et al.  Pedestrian recognition using automotive radar sensors , 2012 .

[10]  Alon Slapak,et al.  Small and Lightweight Innovative Obstacle Detection Radar System for the General Aviation: Performances and Integration Aspects , 2013 .

[11]  Matti Pietikäinen,et al.  Multi-Object Tracking Using Color, Texture and Motion , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[12]  N. H. C. Yung,et al.  A novel method for resolving vehicle occlusion in a monocular traffic-image sequence , 2004, IEEE Transactions on Intelligent Transportation Systems.

[13]  Sharath Pankanti,et al.  Appearance models for occlusion handling , 2006, Image Vis. Comput..