Obtaining optic disc center and pixel region by automatic thresholding methods on morphologically processed fundus images

Marín, D.; Gegúndez-Arias, Manuel Emilio; Suero, Angel; Bravo, José Manuel

doi:10.1016/j.cmpb.2014.11.003

Cited by 95 publications

(39 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…• We use 0 < s 1. This minimum scale due to sampling removes the singularity at (0, 0) from the kernel that solves (43), as proven in [9]. Theorem 1.…”

mentioning

confidence: 85%

“…Lu [51] 99.8% (3) 98.8% (1) 5.0 Lu et al [39] 97.5% (1) 96.3% (3) 40.0 Yu et al [44] 99.1% (11) 4.7 Aquino et al [40] 99.8% (14) 1.7 Giachetti et al [45] 99.7% (4) 5.0 Ramakanth et al [27] 99.4% (7) 100% (0) 93.83% (5) 0.2 Marin et al [43] 99.8% (3) the R 2 templates and the vessel pattern sensitivity of the SE(2) templates. If one of the two ONH characteristics is less obvious (as is e.g.…”

Section: Messidor Drive Stare Time (S)mentioning

confidence: 99%

“…Subsec. 2.4 of the main article) to the single patch problem (43), with (x 0 , θ 0 ) = (0, 0). Here we no longer need a smoothing with a continuous Gaussian G s , as expansion in the B-spline basis already includes regularization.…”

Section: A3 Expansion In B-splinesmentioning

confidence: 99%

“…Average template matching results (± standard deviation) for fovea detection in 5-fold cross validation, number of failed detections in parentheses. 50% ± 3.81% (159) 31.66% ± 9.03% (142) 51.19% ± 5.97% (587) 45.07% ± 3.33% (888) C lin:R 2 45.65% ± 8.61% (113) 98.24% ± 2.74% (3) 98.59% ± 0.36% (17) 91.77% ± 1.26% (133) D lin:R 2 44.21% ± 4.62% (116) 99.49% ± 1.14% (1) 98.84% ± 0.31% (14) 91.90% ± 0.59% (131) E lin:R 2 46.10% ± 8.11% (112) 98.86% ± 1.57% (2) 98.67% ± 0.34% (16) 91.95% ± 1.18% (130) B log:R 2 1.43% ± 1.31% (205) 10.27% ± 5.09% (185) 20.07% ± 3.00% (959) 16.53% ± 2.52% (1349) C log:R 2 9.59% ± 3.74% (188) 70.30% ± 8.57% (61) 77.61% ± 4.64% (267) 68.06% ± 3.53% (516) D log:R 2 11.48% ± 4.70% (184) 83.47% ± 7.80% (32) 88.22% ± 2.81% (141) 77.90% ± 2.00% (357) E log:R 2 2.86% ± 2.62% (202) 79.68% ± 7.92% (40) 84.79% ± 5.16% (181) 73.82% ± 2.62% (423) 19% ± 2.76% (35) 71.53% ± 7.36% (58) 91.31% ± 0.68% (104) 87.81% ± 1.25% (197) C lin:SE(2) 83.65% ± 3.18% (34) 84.13% ± 6.25% (32) 98.23% ± 1.04% (21) 94.62% ± 0.36% (87) D lin:SE(2) 73.57% ± 4.71% (55) 83.69% ± 6.83% (33) 97.88% ± 1.17% (25) 93.01% ± 1.09% (113) E lin:SE(2) 77.83% ± 4.29% (46) 84.88% ± 6.69% (30) 98.22% ± 1.23% (21) 94.00% ± 0.93% (97) B log:SE(2) 75.49% ± 5.73% (51) 60.80% ± 5.68% (80) 92.79% ± 1.98% (86) 86.56% ± 2.20% (217) C log:SE(2) 79.33% ± 6.57%(43) 70.87% ± 10.28%(59) 96.90% ± 0.71% (37) 91.39% ± 1.36% (139) D log:SE(2) 62.09% ± 6.66% (79) 72.57% ± 8.59% (54) 96.64% ± 1.05%(40) 89.30% ± 0.63% (173) E log:SE(2) 68.34% ± 8.59% (66) 72.20% ± 8.53%(55) 96.57% ± 0.96% (41) 89.98% ± 1.25% (162) + C log:SE(2) 96.20% ± 3.15%…”

mentioning

confidence: 99%

See 3 more Smart Citations

Template Matching via Densities on the Roto-Translation Group

Bekkers

Loog

Romeny

et al. 2018

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

We propose a template matching method for the detection of 2D image objects that are characterized by orientation patterns. Our method is based on data representations via orientation scores, which are functions on the space of positions and orientations, and which are obtained via a wavelet-type transform. This new representation allows us to detect orientation patterns in an intuitive and direct way, namely via cross-correlations. Additionally, we propose a generalized linear regression framework for the construction of suitable templates using smoothing splines. Here, it is important to recognize a curved geometry on the position-orientation domain, which we identify with the Lie group SE(2): the roto-translation group. Templates are then optimized in a B-spline basis, and smoothness is defined with respect to the curved geometry. We achieve state-of-the-art results on three different applications: detection of the optic nerve head in the retina (99.83 percent success rate on 1,737 images), of the fovea in the retina (99.32 percent success rate on 1,616 images), and of the pupil in regular camera images (95.86 percent on 1,521 images). The high performance is due to inclusion of both intensity and orientation features with effective geometric priors in the template matching. Moreover, our method is fast due to a cross-correlation based matching approach.

show abstract

“…• We use 0 < s 1. This minimum scale due to sampling removes the singularity at (0, 0) from the kernel that solves (43), as proven in [9]. Theorem 1.…”

mentioning

confidence: 85%

Section: Messidor Drive Stare Time (S)mentioning

confidence: 99%

Section: A3 Expansion In B-splinesmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Template Matching via Densities on the Roto-Translation Group

Bekkers

Loog

Romeny

et al. 2018

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…In most cases, the boundary of the OD and OC are approximated as circular or elliptical shape. In [14], morphological operation and automatic thresholding methods were integrated for the segmentation of OD in fundus images. Firstly, in order to reduce the influence of blood vessels, a series of morphological open and close operations were applied to the original image to obtain a bright regionenhanced image.…”

Section: Introductionmentioning

confidence: 99%

Automatic Segmentation of Optic Disc and Cup in Retinal Fundus Images Using Improved Two‐Layer Level Set Method

Wang

Chi

et al. 2019

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Glaucoma is a group of eye conditions, which can seriously damage optic nerves because of an elevated intraocular pressure. Nowadays, glaucoma has become one of the principal causes of blindness that results in irreversible visual loss. Early screening and treatment of glaucoma can prevent further progression of optic nerve degeneration effectively. e vertical cup-to-disc ratio (CDR) is a commonly used measurement for the detection of glaucoma, and therefore accurate segmentation of optic disc (OD) and optic cup (OC) regions in retinal fundus images is of great significance. In this paper, we present a prior shape constraintbased two-layer level set method for OD and OC segmentation in fundus images. is method uses two different layers of one level set function to represent the OD and OC boundaries. In this method, the distance regularized term is designed to guarantee that the distance between the OD and OC varies smoothly. By introducing the prior shape constraint term energy, the final segmentation results of OD and OC are always in the shape of approximate ellipses. In addition, the proposed method has the property of dealing with the intensity inhomogeneity of fundus images through the local fitting energy embedded. Experiments on images from the Baidu Research database demonstrate that the proposed method has superior performance in terms of accuracy and effectiveness.

show abstract