Evaluation of the Uncertainty of Edge-Detector Algorithms

Anchini, R.; Liguori, Consolatina; Paolillo, Alfredo

doi:10.1109/tim.2007.894909

Cited by 24 publications

(15 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…So edge detection is important to image processing. In references [9] in order to demonstrate the goodness of the edge localization, Rosario Anchini proposed to indicate an interval in which the actual edge falls with a confidence level (CL). As an example, Fig.…”

Section: An Edge Detector Evaluation Method-confidence Levelmentioning

confidence: 99%

“…2 shows, the edge position j 0 of the edge and the interval width Δj, where the edge falls with a specified CL (Δj). [9] Fig. 2 Example of edge-detection uncertainty [9] Fig.…”

Section: An Edge Detector Evaluation Method-confidence Levelmentioning

confidence: 99%

“…Some edge detection evaluation approaches are demonstrated by equations and not suitable to be applied because of the indeterminate weight factors contained in equations. In this paper, a specific confidence level (CL) introduced by Rosario Anchini et al [9] is applied to describe the performance of morphological filter based on parametric edge detection objectively, which is easy to calculate. The values of CL are computed when the images are degraded by speckle with different patterns and different carrier-noise-ratio.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Preliminary investigation of the performance of a ladar image edge detector

Zhou

Yao

et al. 2009

SPIE Proceedings

View full text Add to dashboard Cite

Parametric edge detector has been reported to be successfully applied in actual coherent ladar intensity images corrupted by speckle. But there are isolated erroneous test points in the image processed by parametric test detector. By morphological filter, most isolated missing pixels in edge image background can be removed. Thus, the error of the edge detection can decrease exceedingly which is useful for subsequent image processing. Edge detection plays an important role in ladar image processing. Its capability has direct influence on precision and performance of ladar imaging system. Therefore it is necessary to evaluate the performance of a detector. In order to demonstrate the advantages of morphological filter based on parametric edge detection, the probability that an image pixel may be marked as an edge by this detector is calculated. This detector is applied to detect a large number of images degraded by speckle with different patterns and different carrier-noise-ratio. From the simulation results, the performance of morphological filter based on parametric edge detection is described.

show abstract

Section: An Edge Detector Evaluation Method-confidence Levelmentioning

confidence: 99%

“…2 shows, the edge position j 0 of the edge and the interval width Δj, where the edge falls with a specified CL (Δj). [9] Fig. 2 Example of edge-detection uncertainty [9] Fig.…”

Section: An Edge Detector Evaluation Method-confidence Levelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Preliminary investigation of the performance of a ladar image edge detector

Zhou

Yao

et al. 2009

SPIE Proceedings

View full text Add to dashboard Cite

show abstract

“…• The light intensity of the pixels I(i, j) of the image of the distortion target presents an uncertainty u(I(i, j)) calculated as A Type uncertainty after having captured multiple images of the measurand [25,26]. For calculation purposes, this study assumes that all pixels have the same uncertainty and this takes the maximum value of u(I(i, j)).…”

Section: Model For Calculating Uncertaintiesmentioning

confidence: 99%

Optical measuring equipments. Part I: Calibration model and uncertainty estimation

Caja

Gómez

Maresca

2015

Precision Engineering

View full text Add to dashboard Cite

“…• The luminance of the pixels I(i, j) of the image of the line scale presents an uncertainty u(I), maximum value of u(I(i, j)) calculated as a Type A uncertainty after having captured multiple images of the measured [19].…”

Section: Model For Calculating Uncertaintiesmentioning

confidence: 99%