Edge detection in microscopy images using curvelets

Gebäck, Tobias; Koumoutsakos, Petros

doi:10.1186/1471-2105-10-75

Cited by 67 publications

(36 citation statements)

References 13 publications

(21 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…zero mean Gaussian noise with standard deviation σ, thus producing images with SNR τ /σ. We compare our algorithm (σ = 13, α = 0.5, c = 0.75) to several other algorithms including Matlab implementations of Canny [1] (Smoothing with std 2) and Sobel, Local brightness gradients (PB) [13], Boosted edge learning (BEL) [7], oriented means [3], curvelets [4] and our implementation of beamlets [6]. For evaluation we used the F-measure [14], F = 2P R/(P + R), which trades between precision P and recall R. Figure 4 shows examples of the three patterns along with detection results for the various algorithms.…”

Section: Methodsmentioning

confidence: 99%

“…Anisotropic diffusion methods [2] too face difficulties dealing with low SNRs, as they are typically initialized by local image gradients, whose estimation in noisy images may be unreliable. Recent methods use a variety of filter banks to improve the detection of faint edges, e.g., rectangular filters [3], curvelets [4], shearlets [5], and beamlets [6]. For example, [3,6] use rectangular filters of varying lengths and orientations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Detecting Faint Curved Edges in Noisy Images

Alpert

Galun

Nadler

2010

Computer Vision – ECCV 2010

View full text Add to dashboard Cite

Abstract. A fundamental question for edge detection is how faint an edge can be and still be detected. In this paper we offer a formalism to study this question and subsequently introduce a hierarchical edge detection algorithm designed to detect faint curved edges in noisy images. In our formalism we view edge detection as a search in a space of feasible curves, and derive expressions to characterize the behavior of the optimal detection threshold as a function of curve length and the combinatorics of the search space. We then present an algorithm that efficiently searches for edges through a very large set of curves by hierarchically constructing difference filters that match the curves traced by the sought edges. We demonstrate the utility of our algorithm in simulations and in applications to challenging real images.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Detecting Faint Curved Edges in Noisy Images

Alpert

Galun

Nadler

2010

Computer Vision – ECCV 2010

View full text Add to dashboard Cite

show abstract

“…Finally, the curvelet magnitude image is partitioned into occupied and free regions using a threshold. This approach was first applied for edge detection in microscopy images [10].…”

Section: Methodsmentioning

confidence: 99%

Minimal Correlation Classification

Levy

Wolf

2012

Computer Vision – ECCV 2012

View full text Add to dashboard Cite

Abstract. When the description of the visual data is rich and consists of many features, a classification based on a single model can often be enhanced using an ensemble of models. We suggest a new ensemble learning method that encourages the base classifiers to learn different aspects of the data. Initially, a binary classification algorithm such as Support Vector Machine is applied and its confidence values on the training set are considered. Following the idea that ensemble methods work best when the classification errors of the base classifiers are not related, we serially learn additional classifiers whose output confidences on the training examples are minimally correlated. Finally, these uncorrelated classifiers are assembled using the GentleBoost algorithm. Presented experiments in various visual recognition domains demonstrate the effectiveness of the method.

show abstract

“…Algoritam za detekciju ivica baziran na curvelet transformaciji namenjen slikama dobijenim uz pomoć mikroskopa predložen je u [75], dok je metoda detekcija ivica korišćenjem shearlet transformacije razmatrana u [73].…”

Section: Metode Zasnovane Na Wavelet Transformacijiunclassified

“…Gebäck i Koumoutsakos su modifikovali Canny-jev detektor ivica tako da se umesto određivanja gradijenta i pravca gradijenta na osnovu prvog izvoda Gausovog filtra koristi curvelet transformacija [75]. Na P skala se izračunavaju koeficijenti curvlet transformacije c jlk , gde j označava skalu, l pravac i k = (k 1 ,k 2 ) lokaciju.…”

Section: Metode Zasnovane Na Wavelet Transformacijiunclassified

Wavelet based method for edge detection of defect on coated board during production proccess

Barjaktarović¹

View full text Add to dashboard Cite

RezimeDisertacija razmatra problem detekcije ivica defekata u slikama kartona upotrebom wavelet transformacije. Osobine wavelet transformacije, pre svega predstavljanje singulariteta u signalu malim brojem koeficijenata omogućava realizaciju efikasnog detektora ivica.U okviru ove doktorske disertacije dat je detaljan pregled postojećih detektora ivica baziran na diferenciranju i upotrebi Gausovog filtra. Veoma iscrpno su opisani i postupci zasnovani na wavelet transformaciju. Analizirani su načini poređenja predloženih sa drugim detektorima ivica i veoma su retki slučajevi u kojima autori primenjuju objektivne metode poređenja detektora, a i tada su rezultati komparacije zasnovani na sintetičkoj slici, zbog čega se izvedeni zaključci ne mogu generalizovati na realne slike. Ilustrovani su nedostaci klasičnog detektora ivica kada se primene na slike defekata.Predstavljene su najvažnije karakteristike wavelet transformacije i data je teorijska osnova detekcije singulariteta u signalu upotrebom wavelet transformacije.Kako karakteristike signala imaju uticaj na performanse detektora ivica, ispitivanjem slika defekata na kartonu izveden je matematički model ivice defekata. Na osnovu matematičkog modela i karakteristika slika defekata na kartonu ustanovljeno je da se bolji rezultat ostvaruje kada se u proizvod uključe tri skale wavelet transformacije, a ne samo dve kako je prvobitno predložno, jer je, pre svega, izraženije potiskivanje šuma.Zatim, pokazano je da se predloženi algoritam može primeniti na proizvoljan skup slika, pri čemu se na osnovu karakteristika slika može utvrditi početna skala za formiranje proizvoda koeficijenata wavelet transformacije. U nastavku, prikazane su i analizirane dostupne metodologije komparacije detektora ivica. Ustanovljeno je da se mora primeniti objektivna metoda zasnovana na korišćenju istinite mape ivica. Da bi se ostvarilo adekvatno poređenje, realizovana je baza od 50 slika defekata na kartonu sa odgovarajućim istinitim mapama ivica. Komparacija detektora ivica je izvršena i na osnovu postojeće baze svakodnevnih slika i slika iz vazduha (slike iz ptičje perspektive), kao i odgovarajućih mapa ivica. Za poređenje su izabrani klasični i najčešće korišćeni detektori ivica: Sobel, Canny i Marr-Hildreth, zatim dva detektora bazirana na wavelet transformaciji i noviji, često korišćen detektor ivica -SUSAN (eng. Smallest Univalue Segment Assimilating Nucleus) detektor.Komparacija detektora je pokazala da je predložena metoda detekcije ivica na kartonu superiornija u odnosu na sve navedene detektore ivica. Predloženi detektor daje nešto bolje rezultate i pri poređenju pomoću skupa svakodnevnih slika, kao i skupom slika iz vazduha. Metoda komparacije detektora ivica je proširena sa još dva poređenja.Dodat je šum svakodnevnim slikama kao i slikama iz vazduha i ustanovljeno je da predložena metoda ispoljava evidentno bolje rezultate kod slika sa prisutnim šumom, pri čemu razlika u performansama između predloženog i prvog sledećeg detektora raste kako opada odnos signal-šum.Ova doktorska dis...

show abstract

Edge detection in microscopy images using curvelets

Abstract: Background: Despite significant progress in imaging technologies, the efficient detection of edges and elongated features in images of intracellular and multicellular structures acquired using light or electron microscopy is a challenging and time consuming task in many laboratories.

Cited by 67 publications

References 13 publications

Detecting Faint Curved Edges in Noisy Images

Detecting Faint Curved Edges in Noisy Images

Minimal Correlation Classification

Wavelet based method for edge detection of defect on coated board during production proccess

Contact Info

Product

Resources

About