liu.seSearch for publications in DiVA
Change search
Link to record
Permanent link

Direct link
BETA
Freddie, Åström
Alternative names
Publications (10 of 11) Show all publications
Åström, F., Baravdish, G. & Felsberg, M. (2015). A Tensor Variational Formulation of Gradient Energy Total Variation. In: ENERGY MINIMIZATION METHODS IN COMPUTER VISION AND PATTERN RECOGNITION, EMMCVPR 2015: . Paper presented at 10th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR 2015), 13-16 January 2015, Hong Kong, China (pp. 307-320). Springer Berlin/Heidelberg, 8932
Open this publication in new window or tab >>A Tensor Variational Formulation of Gradient Energy Total Variation
2015 (English)In: ENERGY MINIMIZATION METHODS IN COMPUTER VISION AND PATTERN RECOGNITION, EMMCVPR 2015, Springer Berlin/Heidelberg, 2015, Vol. 8932, p. 307-320Conference paper, Published paper (Refereed)
Abstract [en]

We present a novel variational approach to a tensor-based total variation formulation which is called gradient energy total variation, GETV. We introduce the gradient energy tensor into the GETV and show that the corresponding Euler-Lagrange (E-L) equation is a tensor-based partial differential equation of total variation type. Furthermore, we give a proof which shows that GETV is a convex functional. This approach, in contrast to the commonly used structure tensor, enables a formal derivation of the corresponding E-L equation. Experimental results suggest that GETV compares favourably to other state of the art variational denoising methods such as extended anisotropic diffusion (EAD) and total variation (TV) for gray-scale and colour images.

Place, publisher, year, edition, pages
Springer Berlin/Heidelberg, 2015
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349
National Category
Signal Processing
Identifiers
urn:nbn:se:liu:diva-112270 (URN)10.1007/978-3-319-14612-6_23 (DOI)000357502000023 ()978-3-319-14612-6 (ISBN)978-3-319-14611-9 (ISBN)
Conference
10th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR 2015), 13-16 January 2015, Hong Kong, China
Projects
VIDIVPSEMC^2ETT
Available from: 2014-11-20 Created: 2014-11-20 Last updated: 2018-01-29
Baravdish, G., Svensson, O. & Åström, F. (2015). On Backward p(x)-Parabolic Equations for Image Enhancement. Numerical Functional Analysis and Optimization, 36(2), 147-168
Open this publication in new window or tab >>On Backward p(x)-Parabolic Equations for Image Enhancement
2015 (English)In: Numerical Functional Analysis and Optimization, ISSN 0163-0563, E-ISSN 1532-2467, Vol. 36, no 2, p. 147-168Article in journal (Refereed) Published
Abstract [en]

In this study, we investigate the backward p(x)-parabolic equation as a new methodology to enhance images. We propose a novel iterative regularization procedure for the backward p(x)-parabolic equation based on the nonlinear Landweber method for inverse problems. The proposed scheme can also be extended to the family of iterative regularization methods involving the nonlinear Landweber method. We also investigate the connection between the variable exponent p(x) in the proposed energy functional and the diffusivity function in the corresponding Euler-Lagrange equation. It is well known that the forward problems converges to a constant solution destroying the image. The purpose of the approach of the backward problems is twofold. First, solving the backward problem by a sequence of forward problems we obtain a smooth image which is denoised. Second, by choosing the initial data properly we try to reduce the blurriness of the image. The numerical results for denoising appear to give improvement over standard methods as shown by preliminary results.

Place, publisher, year, edition, pages
Taylor & Francis, 2015
National Category
Mathematical Analysis
Identifiers
urn:nbn:se:liu:diva-111581 (URN)10.1080/01630563.2014.970643 (DOI)000346249200002 ()
Projects
VIDI
Available from: 2014-10-26 Created: 2014-10-26 Last updated: 2017-12-05
Åström, F. & Felsberg, M. (2015). On the Choice of Tensor Estimation for Corner Detection, Optical Flow and Denoising. In: C.V. Jawahar and Shiguang Shan (Ed.), COMPUTER VISION - ACCV 2014 WORKSHOPS, PT II: . Paper presented at Workshop on Emerging Topics in Image Restoration and Enhancement (IREw 2014) in conjunction with Asian Conference on Computer Vision (ACCV 2014) (pp. 16-30). Springer, 9009
Open this publication in new window or tab >>On the Choice of Tensor Estimation for Corner Detection, Optical Flow and Denoising
2015 (English)In: COMPUTER VISION - ACCV 2014 WORKSHOPS, PT II / [ed] C.V. Jawahar and Shiguang Shan, Springer, 2015, Vol. 9009, p. 15p. 16-30Conference paper, Published paper (Refereed)
Abstract [en]

Many image processing methods such as corner detection,optical flow and iterative enhancement make use of image tensors. Generally, these tensors are estimated using the structure tensor. In this work we show that the gradient energy tensor can be used as an alternativeto the structure tensor in several cases. We apply the gradient energy tensor to common image problem applications such as corner detection, optical flow and image enhancement. Our experimental results suggest that the gradient energy tensor enables real-time tensor-based image enhancement using the graphical processing unit (GPU) and we obtain 40% increase of frame rate without loss of image quality.

Place, publisher, year, edition, pages
Springer, 2015. p. 15
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 9009
National Category
Signal Processing
Identifiers
urn:nbn:se:liu:diva-111582 (URN)10.1007/978-3-319-16631-5_2 (DOI)000362451400002 ()9783319166308 (ISBN)9783319166315 (ISBN)
Conference
Workshop on Emerging Topics in Image Restoration and Enhancement (IREw 2014) in conjunction with Asian Conference on Computer Vision (ACCV 2014)
Projects
VIDI
Available from: 2014-10-26 Created: 2014-10-26 Last updated: 2017-04-11Bibliographically approved
Åström, F. (2015). Variational Tensor-Based Models for Image Diffusion in Non-Linear Domains. (Doctoral dissertation). Linköping: Linköping University Electronic Press
Open this publication in new window or tab >>Variational Tensor-Based Models for Image Diffusion in Non-Linear Domains
2015 (English)Doctoral thesis, monograph (Other academic)
Abstract [en]

This dissertation addresses the problem of adaptive image filtering.

Although the topic has a long history in the image processing community, researchers continuously present novel methods to obtain ever better image restoration results.

With an expanding market for individuals who wish to share their everyday life on social media, imaging techniques such as compact cameras and smart phones are important factors. Naturally, every producer of imaging equipment desires to exploit cheap camera components while supplying high quality images. One step in this pipeline is to use sophisticated imaging software including, e.g., noise reduction to reduce manufacturing costs, while maintaining image quality.

This thesis is based on traditional formulations such as isotropic and tensor-based anisotropic diffusion for image denoising. The difference from main-stream denoising methods is that this thesis explores the effects of introducing contextual information as prior knowledge for image denoising into the filtering schemes. To achieve this, the adaptive filtering theory is formulated from an energy minimization standpoint. The core contributions of this work is the introduction of a novel tensor-based functional which unifies and generalises standard diffusion methods. Additionally, the explicit Euler-Lagrange equation is derived which, if solved, yield the stationary point for the minimization problem. Several aspects of the functional are presented in detail which include, but are not limited to, tensor symmetry constraints and convexity. Also, the classical problem of finding a variational formulation to a given tensor-based partial differential equation is studied.

The presented framework is applied in problem formulation that includes non-linear domain transformation, e.g., visualization of medical images.

Additionally, the framework is also used to exploit locally estimated probability density functions or the channel representation to drive the filtering process.

Furthermore, one of the first truly tensor-based formulations of total variation is presented. The key to the formulation is the gradient energy tensor, which does not require spatial regularization of its tensor components. It is shown empirically in several computer vision applications, such as corner detection and optical flow, that the gradient energy tensor is a viable replacement for the commonly used structure tensor. Moreover, the gradient energy tensor is used in the traditional tensor-based anisotropic diffusion scheme. This approach results in significant improvements in computational speed when the scheme is implemented on a graphical processing unit compared to using the commonly used structure tensor.

Place, publisher, year, edition, pages
Linköping: Linköping University Electronic Press, 2015. p. 156
Series
Linköping Studies in Science and Technology. Dissertations, ISSN 0345-7524 ; 1646
Keywords
image diffusion, variational formulation, denoising, tensor, non-linear
National Category
Mathematical Analysis
Identifiers
urn:nbn:se:liu:diva-114279 (URN)10.3384/diss.diva-114279 (DOI)978-91-7519-113-3 (ISBN)
Public defence
2015-03-31, Visionen, Hus B, Campus Valla, Linköpings universitet, Linköping, 09:15 (English)
Opponent
Supervisors
Projects
VIDINACIPGARNICSEMC^2
Available from: 2015-02-20 Created: 2015-02-16 Last updated: 2016-08-31Bibliographically approved
Heinemann, C., Åström, F., Baravdish, G., Krajsek, K., Felsberg, M. & Scharr, H. (2014). Using Channel Representations in Regularization Terms: A Case Study on Image Diffusion. In: Proceedings of the 9th International Conference on Computer Vision Theory and Applications: . Paper presented at 9th International Conference on Computer Vision Theory and Applications (VISAPP 2014), 5-8 January 2014, Lisbon, Portugal (pp. 48-55). SciTePress, 1
Open this publication in new window or tab >>Using Channel Representations in Regularization Terms: A Case Study on Image Diffusion
Show others...
2014 (English)In: Proceedings of the 9th International Conference on Computer Vision Theory and Applications, SciTePress, 2014, Vol. 1, p. 48-55Conference paper, Published paper (Refereed)
Abstract [en]

In this work we propose a novel non-linear diffusion filtering approach for images based on their channel representation. To derive the diffusion update scheme we formulate a novel energy functional using a soft-histogram representation of image pixel neighborhoods obtained from the channel encoding. The resulting Euler-Lagrange equation yields a non-linear robust diffusion scheme with additional weighting terms stemming from the channel representation which steer the diffusion process. We apply this novel energy formulation to image reconstruction problems, showing good performance in the presence of mixtures of Gaussian and impulse-like noise, e.g. missing data. In denoising experiments of common scalar-valued images our approach performs competitive compared to other diffusion schemes as well as state-of-the-art denoising methods for the considered noise types.

Place, publisher, year, edition, pages
SciTePress, 2014
Keywords
Image Enhancement, Channel Representation, Channel Smoothing, Diffusion, Energy Minimization
National Category
Computer Vision and Robotics (Autonomous Systems)
Identifiers
urn:nbn:se:liu:diva-103669 (URN)10.5220/0004667500480055 (DOI)978-989-758-003-1 (ISBN)
Conference
9th International Conference on Computer Vision Theory and Applications (VISAPP 2014), 5-8 January 2014, Lisbon, Portugal
Projects
VIDIGARNICS
Available from: 2014-01-22 Created: 2014-01-22 Last updated: 2018-10-09Bibliographically approved
Åström, F. (2013). A Variational Approach to Image Diffusion in Non-Linear Domains. (Licentiate dissertation). Linköping University Electronic Press
Open this publication in new window or tab >>A Variational Approach to Image Diffusion in Non-Linear Domains
2013 (English)Licentiate thesis, comprehensive summary (Other academic)
Abstract [en]

Image filtering methods are designed to enhance noisy images captured in situations that are problematic for the camera sensor. Such noisy images originate from unfavourable illumination conditions, camera motion, or the desire to use only a low dose of ionising radiation in medical imaging. Therefore, in this thesis work I have investigated the theory of partial differential equations (PDE) to design filtering methods that attempt to remove noise from images. This is achieved by modeling and deriving energy functionals which in turn are minimized to attain a state of minimum energy. This state is obtained by solving the so called Euler-Lagrange equation. An important theoretical contribution of this work is that conditions are put forward determining when a PDE has a corresponding energy functional. This is in particular described in the case of the structure tensor, a commonly used tensor in computer vision.A primary component of this thesis work is to model adaptive image filtering such that any modification of the image is structure preserving, but yet is noise suppressing. In color image filtering this is a particular challenge since artifacts may be introduced at color discontinuities. For this purpose a non-Euclidian color opponent transformation has been analysed and used to separate the standard RGB color space into uncorrelated components.A common approach to achieve adaptive image filtering is to select an edge stopping function from a set of functions that have proven to work well in the past. The purpose of the edge stopping function is to inhibit smoothing of image features that are desired to be retained, such as lines, edges or other application dependent characteristics. Thus, a step from ad-hoc filtering based on experience towards an application-driven filtering is taken, such that only desired image features are processed. This improves what is characterised as visually relevant features, a topic which this thesis covers, in particular for medical imaging.The notion of what are relevant features is a subjective measure may be different from a layman's opinion compared to a professional's. Therefore, we advocate that any image filtering method should yield an improvement not only in numerical measures but also a visual improvement should be experienced by the respective end-user

Place, publisher, year, edition, pages
Linköping University Electronic Press, 2013. p. 32
Series
Linköping Studies in Science and Technology. Thesis, ISSN 0280-7971 ; 1594
National Category
Signal Processing
Identifiers
urn:nbn:se:liu:diva-92788 (URN)LIU-TEK-LIC-2013:28 (Local ID)978-91-7519-606-0 (ISBN)LIU-TEK-LIC-2013:28 (Archive number)LIU-TEK-LIC-2013:28 (OAI)
Presentation
2013-06-13, Visionen, Hus B, Campus Valla, Linköpings universitet, Linköping, 13:15 (English)
Opponent
Supervisors
Projects
NACIP, VIDI, GARNICS
Available from: 2013-05-30 Created: 2013-05-22 Last updated: 2016-05-04Bibliographically approved
Åström, F., Zografos, V. & Felsberg, M. (2013). Density Driven Diffusion. In: 18th Scandinavian Conferences on Image Analysis, 2013: . Paper presented at 18th Scandinavian Conferences on Image Analysis (SCIA 2013), 17-20 June 2013, Espoo, Finland (pp. 718-730).
Open this publication in new window or tab >>Density Driven Diffusion
2013 (English)In: 18th Scandinavian Conferences on Image Analysis, 2013, 2013, p. 718-730Conference paper, Published paper (Refereed)
Abstract [en]

In this work we derive a novel density driven diffusion scheme for image enhancement. Our approach, called D3, is a semi-local method that uses an initial structure-preserving oversegmentation step of the input image.  Because of this, each segment will approximately conform to a homogeneous region in the image, allowing us to easily estimate parameters of the underlying stochastic process thus achieving adaptive non-linear filtering. Our method is capable of producing competitive results when compared to state-of-the-art methods such as non-local means, BM3D and tensor driven diffusion on both color and grayscale images.

Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 7944
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:liu:diva-90016 (URN)10.1007/978-3-642-38886-6_67 (DOI)000342988500067 ()978-3-642-38885-9 (ISBN)978-3-642-38886-6 (ISBN)
Conference
18th Scandinavian Conferences on Image Analysis (SCIA 2013), 17-20 June 2013, Espoo, Finland
Projects
VIDIGARNICSBILDLAB
Available from: 2013-04-08 Created: 2013-03-14 Last updated: 2018-01-24Bibliographically approved
Åström, F., Felsberg, M., Baravdish, G. & Lundström, C. (2013). Targeted Iterative Filtering. In: : . Paper presented at Fourth International Conference on Scale Space and Variational Methods in Computer Vision (SSVM 2013), 2-6 June 2013, Schloss Seggau, Graz region, Austria (pp. 1-11). Springer Berlin/Heidelberg
Open this publication in new window or tab >>Targeted Iterative Filtering
2013 (English)Conference paper, Published paper (Refereed)
Abstract [en]

The assessment of image denoising results depends on the respective application area, i.e. image compression, still-image acquisition, and medical images require entirely different behavior of the applied denoising method. In this paper we propose a novel, nonlinear diffusion scheme that is derived from a linear diffusion process in a value space determined by the application. We show that application-driven linear diffusion in the transformed space compares favorably with existing nonlinear diffusion techniques. 

Place, publisher, year, edition, pages
Springer Berlin/Heidelberg, 2013
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 7893
National Category
Signal Processing
Identifiers
urn:nbn:se:liu:diva-89674 (URN)10.1007/978-3-642-38267-3_1 (DOI)978-3-642-38266-6 (ISBN)978-3-642-38267-3 (ISBN)
Conference
Fourth International Conference on Scale Space and Variational Methods in Computer Vision (SSVM 2013), 2-6 June 2013, Schloss Seggau, Graz region, Austria
Projects
VIDIGARNICSSM10-002BILDLAB
Available from: 2013-04-03 Created: 2013-03-01 Last updated: 2018-01-26Bibliographically approved
Åström, F., Baravdish, G. & Felsberg, M. (2012). On Tensor-Based PDEs and their Corresponding Variational Formulations with Application to Color Image Denoising. In: : . Paper presented at ECCV 2012: 12th European Conference on Computer Vision, 7-12 October, Firenze, Italy (pp. 215-228). Springer Berlin/Heidelberg
Open this publication in new window or tab >>On Tensor-Based PDEs and their Corresponding Variational Formulations with Application to Color Image Denoising
2012 (English)Conference paper, Published paper (Refereed)
Abstract [en]

The case when a partial differential equation (PDE) can be considered as an Euler-Lagrange (E-L) equation of an energy functional, consisting of a data term and a smoothness term is investigated. We show the necessary conditions for a PDE to be the E-L equation for a corresponding functional. This energy functional is applied to a color image denoising problem and it is shown that the method compares favorably to current state-of-the-art color image denoising techniques.

Place, publisher, year, edition, pages
Springer Berlin/Heidelberg, 2012
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; 7574
National Category
Engineering and Technology
Identifiers
urn:nbn:se:liu:diva-79603 (URN)10.1007/978-3-642-33712-3_16 (DOI)978-3-642-33711-6 (ISBN)978-3-642-33712-3 (ISBN)
Conference
ECCV 2012: 12th European Conference on Computer Vision, 7-12 October, Firenze, Italy
Projects
NACIPGARNICSELLIIT
Available from: 2012-08-10 Created: 2012-08-10 Last updated: 2017-06-01Bibliographically approved
Åström, F. & Koker, R. (2011). A parallel neural network approach to prediction of Parkinson´s Disease. Expert systems with applications, 38(10), 12470-12474
Open this publication in new window or tab >>A parallel neural network approach to prediction of Parkinson´s Disease
2011 (English)In: Expert systems with applications, ISSN 0957-4174, E-ISSN 1873-6793, Vol. 38, no 10, p. 12470-12474Article in journal (Refereed) Published
Abstract [en]

Recently the neural network based diagnosis of medical diseases has taken a great deal of attention. In this paper a parallel feed-forward neural network structure is used in the prediction of Parkinson’s Disease. The main idea of this paper is using more than a unique neural network to reduce the possibility of decision with error. The output of each neural network is evaluated by using a rule-based system for the final decision. Another important point in this paper is that during the training process, unlearned data of each neural network is collected and used in the training set of the next neural network. The designed parallel network system significantly increased the robustness of the prediction. A set of nine parallel neural networks yielded an improvement of 8.4% on the prediction of Parkinson’s Disease compared to a single unique network. Furthermore, it is demonstrated that the designed system, to some extent, deals with the problems of imbalanced data sets.

Keywords
Parallel neural networks, Parkinson´s Disease, Decision support system
National Category
Engineering and Technology
Identifiers
urn:nbn:se:liu:diva-69156 (URN)10.1016/j.eswa.2011.04.028 (DOI)
Available from: 2011-06-17 Created: 2011-06-17 Last updated: 2017-12-11
Organisations

Search in DiVA

Show all publications