Academic
Publications
A morphological gradient approach to color edge detection

A morphological gradient approach to color edge detection,10.1109/TIP.2005.864164,IEEE Transactions on Image Processing,Adrian N. Evans,Xin U. Liu

A morphological gradient approach to color edge detection   (Citations: 38)
BibTex | RIS | RefWorks Download
A new color edge detector based on vector differences is proposed. The basic technique gives as its output the maximum distance between the vectors within a mask. When applied to scalar-valued images, the method reduces to the classic mor- phological gradient. The technique is relatively computationally efficient and can also be readily applied to other vector-valued images. To improve the performance in the presence of noise, a novel pairwise outlier rejection scheme is employed. A quantitative evaluation using Pratt's figure of merit shows the new technique to outperform other recently proposed color edge detectors. In addition, application to real images demonstrates the approach to be highly effective despite its low complexity. Index Terms—Color image processing, edge detection, morpho- logical operations.
Journal: IEEE Transactions on Image Processing , vol. 15, no. 6, pp. 1454-1463, 2006
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...[22] proposed a morphological gradient approach, based on the maximum distance between the vectors in a window...
    • ...In each case, the operators are defined as follows: : In our proposed RCMG-Median-Mean estimator, operators and are chosen to be the RCMG gradient vector estimator and the RCMG gradient magnitude estimator, respectively, as follows: For the set of color vectors, let a row (or a column) of a window of size 5 5 be denoted by . is the RCMG operator [22], defined as...
    • ...where denotes the set of pairs of vectors, removed from the set [22]...
    • ... The RCMG operator has been shown to be more powerful than MVD for gradient estimation, when the image is corrupted by salt and pepper noise [22]...
    • ...As stated in [22], the computational complexity of the...
    • ...As color edge detectors, the MVD-Median-Mean, MVD-Median-Max, RCMG-Median-Mean and RCMG-Median-Max estimators are compared with each other and with two other color edge detectors: the MVD edge detector [4] and the RCMG [22]...

    Ehsan Nezhadaryaet al. A New Scheme for Robust Gradient Vector Estimation in Color Images

    • ...In this paper, we present watershed results obtained by the scheme we proposed and described in [15]: First, a one-band robust color morphological gradient (CMG) (RCMG) [26] for the hyperspectral image is computed...
    • ...To overcome the problem of outliers, the RCMG has been proposed [26]...
    • ...If E is a 3 × 3 square structuring element, r =1 is recommended [26]...

    Yuliya Tarabalkaet al. Multiple Spectral–Spatial Classification Approach for Hyperspectral Da...

    • ...Two representative examples of this approach are the multi-color gradient (MCG), proposed by Di Zenzo in [12], and the morphological color gradient of Evans and Liu [13]; the former is used as a benchmark in this paper to compare our algorithm to. While multi-color gradient and related ideas have been used with great success computation of the multi-color gradient for multi-color images with large numbers of bands such as MS and HS imagery ...

    Biliana S. Paskalevaet al. Joint spatio-spectral based edge detection for multispectral infrared ...

    • ...Previous vector-wise color edge detectors such as minimum vector dispersion (MVD) [5] or robust color morphological gradient (RCMG) [6] have been shown to be good gradient magnitude estimators, but can not accurately estimate the gradient direction...
    • ...where Rs is the set of s pairs of vectors removed [6]...
    • ...As an edge detector, the RCMG-Median-Mean operator is compared with two other color edge detectors: MVD edge detector [5] and RCMG [6]...

    Ehsan Nezhadaryaet al. A robust morphological gradient estimator and edge detector for color ...

    • ...In the morphological gradient, we calculated the four gradients as in the traditional method [1][3], and substitute the sum of gradients equation (2), using instead a fuzzy inference system, as we show in Fig. 4...

    O. Mendozaet al. Neural networks recognition rate as index to compare the performance o...

Sort by: