Extraction of line features in a noisy image

Combination of Local and Global Line Extraction

“…Lee and Kweon have proposed a single pass algorithm [17]. The method studies single edge points and links them based on neighboring points and their gradients.…”

A Real-Time Hough-Based Method for Segment Detection in Complex Multisensor Images

“…the standard HT (SHT) [8], the Gerig and Klein (GKHT) method [26], the fast Hough transform (FHT) [18], and image-space methods, e.g. the Burns method [5] and the Lee method [6]. Test images contain only segments (i.e.…”

“…This method allows us to obtain high detection accuracies, but to the detriment of the execution times. Recently, Lee and Kweon [6] developed a faster line extraction algorithm based on the gradient spacepartitioning scheme proposed by Burns et al [5]. This algorithm extracts line segments into a single step without any assumptions and constraints, and with the minimum use of heuristic parameters.…”

Line detection in speckle images using Radon transform and ℓ<inf>1</inf> regularization

“…Lee and Kweon employed Deriche's edge operator, which was claimed to be more robust to noise effects than other operators such as Sobel and Difference of Gaussian filters [2]. Eight directional sticks were employed in a technique by Czerwinski et al [3] that used a rotating kernel transformation to enhance lines and curves in US images.…”