Another look at the dominant point detection of digital curves1☆
References (0)
Cited by (46)
Fast progressive polygonal approximations for online strokes
2023, Graphical ModelsDominant point detection based on suboptimal feature selection methods
2020, Expert Systems with ApplicationsPolygonal approximation based on coarse-grained parallel genetic algorithm
2020, Journal of Visual Communication and Image RepresentationCitation Excerpt :2) Given the error bound, the aim is to obtain the polygon that has the minimum number of vertices, and to make the error between the original shape curve and the approximating polygon not exceed the tolerance of error. Over recent decades, numerous methods have been proposed to solve the two types of polygonal approximation problems, such as (1) Sequential method [1–7]; (2) Spilt method, merge method, and split-and-merge method [8–12]; (3) Dominant and angle detection method [13–25]; (4) Newton’s method [26]; (5) Iteration method [27]; (6) k-means method [28–29]; (7) Dynamic programming [30–33]; (8) Bio-inspired intelligent algorithms [34–39]. Among all these methods, Ray and Ray [2] propose to obtain the polygon of object curve through making the line segment maximize and meanwhile making the approximation error minimize.
Fast computation of optimal polygonal approximations of digital planar closed curves
2016, Graphical ModelsCitation Excerpt :Dominant points are those that can describe the curve for visual perception and recognition. In the literature two major categories can be found: corner detection methods [2–4] and polygonal approximation methods [5–7]. In computer vision polygonal approximation of digital planar curves is an important task for a variety of applications like simplification on vectorization algorithms [8], image analysis [9], shape analysis [10], object recognition [11], Geographical Information Systems [12], and digital cartography [13].
Polygonal approximation of digital planar curves through break point suppression
2010, Pattern RecognitionCitation Excerpt :Those that evaluate the curvature by transforming the contour to the Gaussian scale space [14,35,36,43]. Those that search for dominant points using some significant measure other than curvature [5,9,13,25,19,53,29,55,56,30,7,3,31,15,32,33]. Some of the direct curvature estimation methods use angle detection algorithms and need a previous region of support to estimate the curvature.
- ☆
Electronic Annexes available. See http://www.elsevier.nl/ locate/patrec.