Efficient sampling strategy and refinement strategy for randomized circle detection

Chung, Kuo–Liang; Huang, Yong-Huai; Shi-ming, Shen; Krylov, Andrey S.; Yurin, D. V.; Semeikina, E. V.

doi:10.1016/j.patcog.2011.07.004

Cited by 66 publications

(27 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When C jkl does not contain any curvedness local maximum is discarded. In the last check, it is verified that at least an other edge pixel in V b is close to C jkl , following the four points approach proposed in [10], where it is shown as this allows to reduce the number of necessary iterations until 95%. However, while in [10] the fourth point is randomly sampled from the edge pixel set, in our approach we process all points in V b , exploiting the reduced edge pixel set.…”

Section: Iterative Computation Of the Candidate Circlesmentioning

confidence: 92%

Soccer Ball Detection with Isophotes Curvature Analysis

Marco

Leo

Distante

2013

Image Analysis and Processing – ICIAP 2013

View full text Add to dashboard Cite

Abstract. Circle detection is a critical issue in image analysis: it is undoubtedly a fundamental step in different application contexts, among them one of the most challenging is the detection of the ball in soccer game. Hough Transform based circle detector are largely used but there is a large open research area that attempt to provide more effective and less computationally expensive solutions based on randomized approaches, i.e. based on iterative sampling of the edge pixels. To this end, this work presents an ad-hoc randomized iterative work-flow, which exploits geometrical properties of isophotes, the curvature, to identify edge pixels belonging to the ball boundaries; this allow to consider a large amount of edge pixels, but limiting most of the time-consuming computation only on a restricted subset given by pixels with an high probability to lie on a circular structure. The method, coupled with a background suppression algorithm, has been applied to a set of real images acquired by fixed camera providing performances higher than a standard circular Hough transform solver, with a detection rate > 86%.

show abstract

Section: Iterative Computation Of the Candidate Circlesmentioning

confidence: 92%

Soccer Ball Detection with Isophotes Curvature Analysis

Marco

Leo

Distante

2013

Image Analysis and Processing – ICIAP 2013

View full text Add to dashboard Cite

show abstract

“…Triangle detectors are mainly applied to traffic sign recognition using the Hough transform to detect lines of the triangles [19,24] . Circle detectors [14,15] are used for natural images, medical images and document graphs. Rectangle detectors extract parallel and perpendicular contour segments to find rectangles [25] .…”

Section: Geometric Shape Analysis and Recognitionmentioning

confidence: 99%

Improving retrieval of plane geometry figure with learning to rank

Liu

Liao

et al. 2016

Pattern Recognition Letters

View full text Add to dashboard Cite

“…Thus, a number of previous studies aimed at reducing the storage/time requirements of the SCHT (e.g. Chiu and Liaw, 2005;Guo et al, 2006;Chung et al, 2012;Huang et al, 2012). Convolution filters for SCHT and the invariance kernels were also proposed as alternative approaches for the search performed in the parameter space (e.g.…”

Section: Previous Studiesmentioning

confidence: 99%

Automated Detection of Oil Depots From High Resolution Images: A New Perspective

Baseski

2015

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:This paper presents an original approach to identify oil depots from single high resolution aerial/satellite images in an automated manner. The new approach considers the symmetric nature of circular oil depots, and it computes the radial symmetry in a unique way. An automated thresholding method to focus on circular regions and a new measure to verify circles are proposed. Experiments are performed on six GeoEye-1 test images. Besides, we perform tests on 16 Google Earth images of an industrial test site acquired in a time series manner (between the years 1995 and 2012). The results reveal that our approach is capable of detecting circle objects in very different/difficult images. We computed an overall performance of 95.8% for the GeoEye-1 dataset. The time series investigation reveals that our approach is robust enough to locate oil depots in industrial environments under varying illumination and environmental conditions. The overall performance is computed as 89.4% for the Google Earth dataset, and this result secures the success of our approach compared to a state-of-the-art approach.

show abstract

Efficient sampling strategy and refinement strategy for randomized circle detection

Cited by 66 publications

References 19 publications

Soccer Ball Detection with Isophotes Curvature Analysis

Soccer Ball Detection with Isophotes Curvature Analysis

Improving retrieval of plane geometry figure with learning to rank

Automated Detection of Oil Depots From High Resolution Images: A New Perspective

Contact Info

Product

Resources

About