Automatic Recognition of Color Fundus Photographs

Yamamoto, Shozo; Yokouchi, Hisatake

doi:10.1007/978-1-4684-0769-3_24

Cited by 13 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Earliest attempts of determining the characteristics of fundus images can be traced back to 1970s (Pringle 1969;Yamamoto and Yokouchi 1976;Akita and Kuga 1982;Tanaka and Tanaka 1980;Okamoto et al 1988); however, a concrete work on retinal vessel detection from fundus image using 2D-matched filters was proposed by Chaudhuri et al (1989). Their proposed algorithm assumes that (a) vessels can be segmented using piece-wise linear segments, (b) gray level intensity distribution of a vessel can be approximated with a Gaussian curve, and (c) width of a vessel is constant.…”

Section: Introductionmentioning

confidence: 99%

High speed detection of retinal blood vessels in fundus image using phase congruency

Amin

Yan

2010

Soft Comput

View full text Add to dashboard Cite

Detection of blood vessels in retinal fundus image is the preliminary step to diagnose several retinal diseases. There exist several methods to automatically detect blood vessels from retinal image with the aid of different computational methods. However, all these methods require lengthy processing time. The method proposed here acquires binary vessels from a RGB retinal fundus image in almost real time. Initially, the phase congruency of a retinal image is generated, which is a softclassification of blood vessels. Phase congruency is a dimensionless quantity that is invariant to changes in image brightness or contrast; hence, it provides an absolute measure of the significance of feature points. This experiment acquires phase congruency of an image using LogGabor wavelets. To acquire a binary segmentation, thresholds are applied on the phase congruency image. The process of determining the best threshold value is based on area under the relative operating characteristic (ROC) curve. The proposed method is able to detect blood vessels in a retinal fundus image within 10 s on a PC with (accuracy, area under ROC curve) = (0.91, 0.92), and (0.92, 0.94) for the STARE and the DRIVE databases, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

High speed detection of retinal blood vessels in fundus image using phase congruency

Amin

Yan

2010

Soft Comput

View full text Add to dashboard Cite

show abstract

“…The corresponding points are projected onto the surface and in 3D space we transform one of the image pair using these projected point pairs. Second, we define the error function J = : Q -plkI2 (2) P'k RPk+T (3) where Pk is the the coordinate of the kth corresponding point on the surface re-constructed by the reference image, P1k S the coordinate of P,, after the geometric transformation (Equation 3), Qk is the coordinate of the kth corresponding point on the non-diagnosed image, R and T are variables defining the rotation matrix and the translation vector, respectively. This transformation can match a pair of images everywhere, while linear (affine) transformation matches them only in the neighborhood of the corresponding points.…”

Section: Matching the Position Of Two Imagesmentioning

confidence: 99%

<title>Automated detection of changes in sequential color ocular fundus images</title>

et al. 1998

View full text Add to dashboard Cite

A recent trend is the automatic screening of color ocular fundus images. The examination of such images is used in the early detection of several adult diseases such as hypertension and diabetes. Since this type of examination is easier than CT, costs less, and has no harmful side effects, it will become a routine medical examination. Normal ocular fundus images are found in more than 90% of all people. To deal with the increasing number of such images, this paper proposes a new approach to process them automatically and accurately. Our approach, based on individual comparison, identifies changes in sequential images: a previously diagnosed normal reference image is compared to a non-diagnosed image.This paper presents a new approach for measuring changes in sequential color ocular fundus images and presents test results that prove its detection efficiency. Two processes that are required are registration and the detection of color differences in the two images. First, the images are registered based on the individual's ocular fundus structure. Second, the images are divided into blocks and for each pair of blocks, the chromaticity and luminosity are extracted. This block-wise division offsets the effect of environmental variations which extend over much larger areas than the blocks. If one block diverges sufficiently from its reference, it signifies a change in the block.The initial screening results, 350 image pairs, showed that 86% of the pairs (50 abnormal and 300 normal image pairs) had no significant difference and were successfully detected as normal with no false negatives. This approach suits the processing of image pairs where there is a need to detect small changes accurately regardless of environmental changes. Prior to clinical use, the proposed approach must be more extensively evaluated using a wide sample.

show abstract

“…Early detection and treatment of these diseases are crucial to avoid this preventable vision loss. Computer based retinal image analysis has been worked out since 1974 [1] and becoming very popular for quick detection of retinal diseases.…”

Section: Introductionmentioning

confidence: 99%

Automated detection of optic disc and blood vessel in retinal image using morphological, edge detection and feature extraction technique

Mithun¹,

Das²,

Fattah

2014

16th Int'l Conf. Computer and Information Technology

View full text Add to dashboard Cite

Reliable, fast and efficient optic disc localization and blood-vessel detection are the primary tasks in computer analyses of retinal image. Most of the existing algorithms suffer due to inconsistent image contrast, varying individual condition, noises and computational complexity. This paper presents an algorithm to automatically detect landmark features of retinal image, such as optic disc and blood vessel. First, optic disc and blood vessel pixels are detected from blue plane of the image. Then, using OD location the vessel pixels are connected. The detection scheme utilizes basic operations like edge detection, binary thresholding and morphological operation. This method was evaluated on standard retinal image databases, such as STARE and DRIVE. Experimental results demonstrate that the high accuracy achieved by the proposed method is comparable to that reported by the most accurate methods in literature, alongside a substantial reduction of execution time. Thus the method may provide a reliable solution in automatic mass screening and diagnosis of the retinal diseases.

show abstract

Automatic Recognition of Color Fundus Photographs

Cited by 13 publications

References 0 publications

High speed detection of retinal blood vessels in fundus image using phase congruency

High speed detection of retinal blood vessels in fundus image using phase congruency

<title>Automated detection of changes in sequential color ocular fundus images</title>

Automated detection of optic disc and blood vessel in retinal image using morphological, edge detection and feature extraction technique

Contact Info

Product

Resources

About