A survey of subpixel edge detection methods for images of heat-emitting metal specimens

Abstract

In this paper the problem of accurate edge detection in images of heat-emitting specimens of metals is discussed. The images are provided by the computerized system for high temperature measurements of surface properties of metals and alloys. Subpixel edge detection is applied in the system considered in order to improve the accuracy of surface tension determination. A reconstructive method for subpixel edge detection is introduced. The method uses a Gaussian function in order to reconstruct the gradient function in the neighborhood of a coarse edge and to determine its subpixel position. Results of applying the proposed method in the measurement system considered are presented and compared with those obtained using different methods for subpixel edge detection.

Sub-pixel Edge Detection of LED Probes Based On Partial Area Effect

“…To increase the precision of edge detection, several subpixel techniques had been proposed, including curve-fitting, moment-based [3][4][5][6], reconstructive [7] and partial area effect methods [8]. These methods have their characteristics.…”

Edge reconstruction of LED probes using various segmentation and the averaging of sub-pixels

“…While the right and left edge line of probes are both almost perpendicular to the x-axis, reconstruction is demonstrated focusing on horizontal direction. Then establish a model by using curve-fitting method to find the maximum value of gradient curve according to that specific region, the position of this value presents the sub-pixel edge point (refer to [2]). After gaining the edge points of sub-pixels, generally these points are used directly for angle calculation by iterative curve fitting.…”

“…Our main target is to find methods for enhancement on precision and efficiency. precision in edge detection effectively, including reconstructive [2], partial area effect [3][4], and moment-based methods [5][6][7][8]. Tabatabaic and Mitchell were the first two person to use sub-pixel technology to increase precision of edge detection, this method can overcome measurement error on edge caused by the effect from noise.…”

Cross-task code reuse in genetic programming applied to visual learning

“…Note that our feature extraction method is simple and could be improved by applying more accurate edge detection algorithms as described by, e.g., Fabijańska (2012), or by using a more sophisticated interest point detector, proposed, e.g., by Trujillo and Olague (2006). Each primitie is described by three scalars called hereafter attributes; these include two spatial coordinates of the edge fragment (x and y) and the local gradient orientation.…”