Approach for transient interference detection based on straight line extraction for high-frequency sky-wave radar

Transient interference cancellation is necessary for high-frequency sky-wave radar and usually requires accurate interference detection and localisation which is complicated due to the non-uniform noise level caused by sea-clutter mitigation in Doppler domain. A novel approach is proposed by converting the detection of interference into a problem of straight line extraction in the constructed fast- and slow-time images. The line edges are detected by an edge detector, such as Canny method, followed by Hough transform. Then, each edge is classified as a rising, falling or false-alarm edge based on multi-threshold segmentation of image sub-blocks. Experimental results demonstrate the effectiveness of the proposed approach.