Recovering 3D Shape of Weak Textured Surfaces

Muhammad, Mannan Saeed; Mutahira, Husna; Majid, Abdul; Choi, Tae‐Sun

doi:10.1109/iccsa.2009.25

Cited by 7 publications

(5 citation statements)

References 8 publications

(10 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…[5]), over differently shaped kernels, to nonlinear filtering as proposed in [6]. In [7,8], the authors proposed to detect regions with a high variance in the depth map and suggested to smooth the depth in these parts by interpolation. Pertuz, Puig, and Garcia analyzed the behavior of the contrast curve in order to identify and remove low-reliability regions of the depth map in [9].…”

Section: Introductionmentioning

confidence: 99%

Variational Depth From Focus Reconstruction

Moeller¹,

Benning²,

Schönlieb³

et al. 2015

IEEE Trans. on Image Process.

100

View full text Add to dashboard Cite

This paper deals with the problem of reconstructing a depth map from a sequence of differently focused images, also known as depth from focus (DFF) or shape from focus. We propose to state the DFF problem as a variational problem, including a smooth but nonconvex data fidelity term and a convex nonsmooth regularization, which makes the method robust to noise and leads to more realistic depth maps. In addition, we propose to solve the nonconvex minimization problem with a linearized alternating directions method of multipliers, allowing to minimize the energy very efficiently. A numerical comparison to classical methods on simulated as well as on real data is presented.

show abstract

Section: Introductionmentioning

confidence: 99%

Variational Depth From Focus Reconstruction

Moeller¹,

Benning²,

Schönlieb³

et al. 2015

IEEE Trans. on Image Process.

100

View full text Add to dashboard Cite

show abstract

“…Several algorithms and operators have been proposed for this purpose. The most popular operators are the Modified Laplacian [8], the Tenengrad Algorithm [23] and the GrayLevel Variance [20], among others [24]. In the present work, the Modified Laplacian is used as the focus measure:…”

Section: Focus Measurementioning

confidence: 99%

“…Adaboost is then used to combine each weak classifier in order to obtain the best classification rate in the training set. 2 In the experiments, a total of five filtering algorithms have been used in the comparisons: the Canny edge detector-based algorithm proposed in [22] (CAN), the mean response of the Gabor filters (GAB), the combination of Gabor filters using Adaboost (G+AD), the depth-map filtering-based algorithm proposed in [20] (DFIL) and the proposed reliability-based method (R2). In order to compare the effect of only the filtering stage of these approaches, all the depth-maps have been computed with the same SFF algorithm [7].…”

Section: Accepted M Manuscriptmentioning

confidence: 99%

See 1 more Smart Citation

Reliability measure for shape-from-focus

Pertuz

Puig

García

2013

Image and Vision Computing

View full text Add to dashboard Cite

This is the author’s version of a work that was accepted for publication in Journal Image and Vision Computing . Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Journal Image and Vision Computing , 31, 10 (2013) DOI: 10.1016/j.imavis.2013.07.005Shape-from-focus (SFF) is a passive technique widely used in image processing for obtaining depth-maps. This technique is attractive since it only requires a single monocular camera with focus control, thus avoiding correspondence problems typically found in stereo, as well as more expensive capturing devices. However, one of its main drawbacks is its poor performance when the change in the focus level is difficult to detect. Most research in SFF has focused on improving the accuracy of the depth estimation. Less attention has been paid to the problem of providing quality measures in order to predict the performance of SFF without prior knowledge of the recovered scene. This paper proposes a reliability measure aimed at assessing the quality of the depth-map obtained using SFF. The proposed reliability measure (the R-measure) analyses the shape of the focus measure function and estimates the likelihood of obtaining an accurate depth estimation without any previous knowledge of the recovered scene. The proposed R-measure is then applied for determining the image regions where SFF will not perform correctly in order to discard them. Experiments with both synthetic and real scenes are presented

show abstract

“…Amb aquest mètode, es basa en detectar la textura present de la mostra que es vol mesurar. Amb aquest mètode, es captura una imatge de camp clar, i un algoritme anomenat operador de focus (Focus Operator) [23] és capaç d'extreure la posició axial del focus per a cada píxel de la imatge. Quan la superfície està en focus la seva textura està ben definida en la imatge i l'operador detecta un senyal elevat, mentre que si la superfície està fora de focus, la imatge està borrosa i el senyal detectat per l'operador es redueix dràsticament.…”

Section: Texturització Artificialunclassified

Estudi i desenvolupament de sensors òptics ultra-ràpids per a metrologia tridimensional de superfícies

Martínez Marín

View full text Add to dashboard Cite

(English) Three-dimensional measurement of roughness and shape of a surface provides information about an object properties through surface parameters. These parameters are defined by international standards and can describe the feature behaviour of its surface. The main benefit of using optical measurement techniques is that the sample under inspection does not suffer any damage when measured. With contact-based methods, otherwise, the damages can occur. With the evolution towards industry 4.0, there is a need to measure faster with smaller sensors, in order to have a manufacture-quality control close loop. This leads to the objective of this thesis, which is to develop two ultrafast sensors for 3d surface metrology. The first sensor consists in an evolution of an existing commercial system using conventional techniques. It has to be reduced in dimensions and cost-wise, while increasing its acquisition speed. The second sensor must use a single-shot technique, so that with one single image, the sample's surface's heights can be retrieved. Therefore, the current optical techniques for 3d surface metrology are reviewed. There are classified in three-main categories: classical techniques, single-shot techniques and scanning systems without moving parts. The development of the thesis is divided into two parts, corresponding to each sensor. In the first part, the opto-mechanical design of the sensor is developed, together with its corresponding analysis. One of the most relevant changes is the replacement of the microdisplay with a glass checkerboard pattern. The microdisplay is a critical component for measuring with confocal, so the algorithms have had to be adapted to obtain equivalent results. In addition, sinusoidal movement profiles have been studied to reduce the sensor's downtime and minimize motor wear. In the second part the objective is to demonstrate the feasibility of using CKM (Complementary Kernel Matching) for surface measurement. This technique is based on the image processing after a calibration that analyzes the Point Spread Function (PSF) through the optical axis. In order to demonstrate it, a prototype based on two cameras has been developed, a new calibration method has been designed, which divides the image into smaller regions using a matrix of pinholes. A new algorithm has also been developed to detect the height of the sample from the axial response obtained. Finally, a real sample has been measured with satisfactory results. (Català) La mesura tridimensional de forma i rugositat de textura d’una superfície permet proporcionar informació sobre les propietats d’un objecte a partir d'un conjunt de paràmetres de superfície. Aquests paràmetres estan definits per normatives internacionals i poden descriure directament el comportament de la funcionalitat de la seva superfície. El principal benefici d’utilitzar tècniques de mesura òptiques és que la mostra sota inspecció no pateix danys ni desperfectes a l’hora de ser mesurada, tal i com sí pot succeir en els mètodes de contacte. En el procés actual d’evolució cap a la indústria 4.0, es necessita mesurar més ràpid amb sensors més petits, per tal de tenir un circuit tancat de fabricació-control de qualitat. És per això que l'objectiu d'aquesta tesi és desenvolupar dos sensors ultra-ràpids per a la mesura tridimensional de superfícies. El primer consisteix en evolucionar un equip comercial amb tècniques convencionals en els qual s'ha de reduir les dimensions i els costos dels components, a la vegada que s'augmenta la velocitat d'adquisició. El segon sensor ha de funcionar amb una tècnica single-shot, de manera que amb l'adquisició d'una sola imatge, el sensor ja recuperi les alçades de la superfície de la mostra. En conseqüència, en la tesi es revisen les actuals tècniques de mesura òptiques per a metrologia tridimensional de superfícies. Aquestes es divideixen en tres blocs: tècniques clàssiques, tècniques d'una sola imatge i sistemes d'escombrat sense components mecànics. El desenvolupament de la tesi es divideix en dues parts, que corresponen a cada sensor. En la primera part es desenvolupa el disseny opto-mecànic del sensor, amb el seu anàlisis corresponent. Un dels canvis més rellevants és la substitució del microdisplay per un patró checkerboard de vidre. El microdisplay és un component crític per a poder mesurar amb confocal, de manera que s'han hagut de modificar els algoritmes per a poder obtenir resultats equivalents. Addicionalment s'han estudiat perfils de moviment sinusoïdals per a milllorar el temps productiu del sensor i minimitzar el desgast del motor. En la segona part es vol demostrar la viabilitat conceptual d'emprar CKM (Complementary Kernel Matching) per a la mesura de superfícies. Aquesta tècnica es basa en el processat d'imatge després d'una calibració que analitza la resposta impulsional (PSF) a través de l'eix òptic. Per a demostrar-la, s'ha desenvolupat un prototip basat en dues càmeres, s'ha dissenyat un nou mètode de calibració basat en la divisió de la imatge en finestres gràcies a utilitzar una matriu de pinholes. També s'ha desenvolupat un nou algoritme per a detectar l'alçada de la mostra a partir de la resposta axial obtinguda. Finalment, s'ha mesurat una mostra real amb resultats satisfactoris.

show abstract

Recovering 3D Shape of Weak Textured Surfaces

Cited by 7 publications

References 8 publications

Variational Depth From Focus Reconstruction

Variational Depth From Focus Reconstruction

Reliability measure for shape-from-focus

Estudi i desenvolupament de sensors òptics ultra-ràpids per a metrologia tridimensional de superfícies

Contact Info

Product

Resources

About