Research interests


Distance measures between objects and images

We study distance measures between sets which can be applied in image processing. We have proposed new distance measures and have evaluated their performance in template matching, object classification, and image registration. We are particularly interested in distance which are applicable to information-rich object representations, conveniently modeled by fuzzy sets.

Our latest result, a proposed image registration framework for symmetric, intensity-interpolation free, precise, affine image registration, exhibits excellent performance. Code is available here.

People: Joakim Lindblad, Johan Öfverstedt, Vladimir Ćurić, Gunilla Borgefors

Published results:

  • J. Öfverstedt, J. Lindblad, N. Sladoje. Stochastic Distance Transform. In Proceedings of the 21th International Conference on Discrete Geometry for Computer Imagery, DGCI2019, Paris, France, March 2019. Accepted.
  • J. Öfverstedt, J. Lindblad, N. Sladoje. Fast and Robust Symmetric Image Registration Based on Distances Combining Intensity and Spatial Information. IEEE Transactions on Image Processing, Feb. 2019.
  • N. Sladoje and J. Lindblad. Distance Between Vector-valued Representations of Objects in Images with Application in Object Detection and Classification. In Proc. of the 18th International Workshop on Combinatorial Image Analysis, IWCIA2017, Plovdiv, Bulgaria, Lecture Notes in Computer Science, LNCS-10256, pp. 243-255, 2017.
  • J. Öfverstedt, N. Sladoje, and J. Lindblad. Distance Between Vector-valued Fuzzy Sets based on Intersection Decomposition with Applications in Object Detection. In Proc. of the 13th International Symposium on Mathematical Morphology, ISMM2017, Fontainebleau, France, Lecture Notes in Computer Science, LNCS-10225, pp. 395-407, 2017.
  • J. Lindblad and N. Sladoje. Exact Linear Time Euclidean Distance Transforms of Grid Line Sampled Shapes. In Proc. of the 12th Intern. Symposium on Mathematical Morphology (ISMM), Reykjavik, Iceland. Lecture Notes in Computer Science, Vol. 9082, pp. 645-656, Springer, 2015.
  • V. Ćurić, J. Lindblad, N. Sladoje, H. Sarve, and G. Borgefors. A new set distance and its application to shape registration. Pattern Analysis and Applications, Vol. 17, No. 1, pp. 141-152, 2014.
  • J. Lindblad and N. Sladoje. Linear time distances between fuzzy sets with applications to pattern matching and classification. IEEE Transactions on Image Processing, Vol. 23, No 1, pp. 126-136, 2014.
  • V. Ćurić, J. Lindblad, and N. Sladoje. Distance measures between digital fuzzy objects and their applicability in image processing. In Proc. of the 14th International Workshop on Combinatorial Image Analysis (IWCIA2011), Madrid, Spain. Lecture Notes in Computer Science, Vol. 6636, pp. 385-395 , 2011.
  • J. Lindblad, V. Ćurić, and N. Sladoje. On set distances and their application to image registration. In Proc. of the 6th International Symposium on Image and Signal Processing and Analysis (ISPA),Salzburg, Austria. IEEE, pp. 449-454, 2009.


Image Restoration Based on Energy Minimization

A common approach to solve the very important but severely ill-posed problem of image deconvolution, is to formulate it in a form of an energy minimization problem. We study performances of energy minimization based restoration methods for enhancing images degraded with blur and different types of noise types - Gaussian, Poisson and mixed Poisson-Gaussian.

People: Buda Bajić, Joakim Lindblad, Tibor Lukić

Published results:

  • B. Bajić, J. Lindblad, N. Sladoje. Sparsity promoting super-resolution coverage segmentation by linear unmixing in presence of blur and noise. J. Electron. Imaging 28(1), 013046, 2019.
  • B. Bajić, A. Suveer, A. Gupta, I. Pepić, J.Lindblad, N. Sladoje, I.-M. Sintorn. Denoising of Short Exposure Transmission Electron Microscopy Images for Ultrastructural Enhancement. In Proc. of the IEEE International Symposium on Biomedical Imaging, ISBI2018, Washington, D.C., USA, pp.921-925, 2018.
  • B. Bajić, J. Lindblad, N. Sladoje. Single image super-resolution reconstruction in presence of mixed Poisson-Gaussian noise. Accepted for the 6th International Conference on Image Processing Theory, Tools and Applications, IPTA 2016, Oulu, Finland.
  • B. Bajić, J. Lindblad, N. Sladoje. Restoration of images degraded with signal-dependent noise based on energy minimisation: an empirical study. Journal of Electronic Imaging. 25(4), 043020 2016.
  • B. Bajić, J. Lindblad, N. Sladoje. Blind Restoration of Images Degraded with Mixed Poisson-Gaussian Noise with Application in Transmission Electron Microscopy. In Proc. of 2016 IEEE Intern. Symposium on Biomedical Imaging, ISBI2016, Prague, Czech Republic, pp. 123-127, 2016.
  • B. Bajić, J. Lindblad, and N. Sladoje. An Evaluation of Potential Functions for Regularized Image Deblurring. In Proceedings of the International Conference on Image Analysis and Recognition (ICIAR), Algarve, Portugal. Lecture Notes in Computer Science, Vol. 8814, pp. 150–158, 2014.
  • T. Lukić, J. Lindblad, N. Sladoje. Regularized image denoising based on spectral gradient optimization. Inverse Problems, Vol. 27, No. 8, 085010, 2011.
  • J. Lindblad, N. Sladoje, and T Lukić. De-noising of SRµCT Fiber Images by Total Variation Minimization. In Proc. of the 20th International Conference on Pattern Recognition (ICPR),Istanbul, Turkey. IEEE, pp. 4621-4624, 2010.


Multi-layer representations for texture analysis in biomedical image processing

We are developing the theoretical foundation for a class of methods applicable to multi-layered heterogeneous object representations. Focus has been on texture descriptors applied to multi-scale data to allow for a search of candidate areas that are likely to hold objects of interest in low resolution images. Convolutional neural networks are, in a number of different ways, combined with information extracted from Local Binary Pattern features, to provide a powerful tool for texture-based classification of biomedical data.

People: Elisabeth Wetzer, Joakim Lindblad, Ida-Maria Sintorn, Kjell Hultenby, Jo Gay, Hugo Harlin

Published results:

  • E. Wetzer, J. Lindblad, I.-M. Sintorn, K. Hultenby, N. Sladoje. Towards automated multiscale imaging and analysis in TEM: Glomerulus detection by fusion of CNN and LBP maps.In Proc. of the ECCV 2018, Workshop on BioImage Computing, Lecture Notes in Computer Science, LNCS-11134, pp. 465-475, Munich, Germany, Sept. 2018.
  • J. Gay, H. Harlin, E. Wetzer, J. Lindblad, and N. Sladoje. Texture-based oral cancer detection: A performance analysis of deep learning approaches. Poster Δ at the 3rd NEUBIAS Conference, Luxembourg, February 2019.
  • E. Wetzer, J. Lindblad, I-M. Sintorn, K. Hultenby, and N. Sladoje. Towards automated multiscale Glomeruli detection and analysis in TEM by fusion of CNN and LBP maps. Poster Δ at the 3rd NEUBIAS Conference, Luxembourg, February 2019.


Robust learning of geometric equivariances

We build on recent works on Geometric deep learning and aim at combining it with Manifold learning, to produce truly learned equivariances without the need for engineered solutions and maximize benefits of shared weights (parameters to learn). A decrease of the number of parameters to learn leads to increased performance, generalizability and reliability (robustness) of the network. An additional gain is in reducing a risk that the augmented data incorporates artefacts not present it the original data; this is particularly relevant for reliable texture-based analysis and classification in biomedical applications.

This project is conducted within AI-Math track of WASP – the Wallenberg Artificial Intelligence, Autonomous Systems and Software Program, a major Swedish national initiative for strategically motivated basic research, education and faculty recruitment.

People: Karl Bengtsson Bernander, Joakim Lindblad


Explainable AI - Visualization of CNN class activations in cancer detection

We aim to increase understanding and interpretability of the classification behaviour of deep networks by visualizing relevant steps in the process. In particular, we work towards understanding what parts of cell images are considered as most important for prediction when trained deep CNNs are used to differentiate cytological images into normal and abnormal classes.

People: Nadezhda Koriakina, Joakim Lindblad, Ewert Bengtsson, Eva Darai Ramqvist, Jan-Michaél Hirsch, Christina Runow Stark

Published results:

  • N. Koriakina, N. Sladoje, E. Bengtsson, E. Darai Ramqvist, J-M. Hirsch, C. Runow Stark, J. Lindblad. Visualization of convolutional neural network class activations in automated oral cancer detection for interpretation of malignancy associated changes. Poster Δ at the 3rd NEUBIAS Conference, Luxembourg, February 2019.


Coverage Model and its Application to High Precision Image Processing

The coverage model, which we have been developing for several years now, provides a framework for representing objects present in digital images as spatial fuzzy subsets. Assigned membership values indicate to what extent image elements are covered by the imaged objects. During last years, we have shown, both theoretically, and in applications, that the model can be used to improve information extraction from digital images and to reduce problems originating from limited spatial resolution.

People Joakim Lindblad, Slobodan Dražić, Vladimir Ilić, Kristina Lidayova, Zoltan Kato, Attila Tanács

Published results:

  • S. Dražić, N. Sladoje, and J. Lindblad. Accurate Estimation of Feret’s diameter of a Shape from Pixel Coverage Digitization. Pattern Recognition Letters, Vol.80, pp. 37–45, 2016.
  • V. Ilić, J. Lindblad and N. Sladoje. Signature of a shape based on its pixel coverage representation. In Proc. of the 19th international conference on Discrete Geometry for Computer Imagery, DGCI 2016, Nantes, France. Lecture Notes in Computer Science, Vol. 9647, pp. 181-193, 2016.
  • V. Ilić, J. Lindblad and N. Sladoje. Precise Euclidean distance transforms in 3D from voxel coverage representation. Pattern Recognition Letters, Vol. 65(1), pp. 184-191, 2015.
  • K. Lidayova, J.Lindblad, N. Sladoje, H.Frimmel, C.Wang, and Ö.Smedby. Coverage Segmentation of 3D Thin Structures. In Proc. of the 5th International Conference on Image Processing Theory, Tools and Applications, IPTA 2015, Orleans, France, pp. 23-28, 2015.
  • K. Lidayova, J. Lindblad, N. Sladoje and H. Frimmel. Coverage segmentation of thin structures by linear unmixing and local centre of gravity attraction. In Proc. of the 8th IEEE International Symposium on Image and Signal Processing and Analysis (ISPA 2013), Trieste, Italy, 6046670, pp.83-88, 2013.
  • N. Sladoje and J. Lindblad. The coverage model and its use in image processing. Book chapter in: Selected Topics on Image Processing and Cryptology (Ed. Miodrag Mihaljević), Zbornik radova (Collection of Papers), No 15(23), pp. 39-117, Mathematical Institute of the Serbian Academy of Sciences and Arts, Belgrade, 2012.
  • J. Lindblad and N. Sladoje. Coverage Segmentation based on Linear Unmixing and Minimization of Perimeter and Boundary Thickness. Pattern Recognition Letters, Vol. 33, No.6, pp. 728-738, 2012.
  • S. Dražić, J. Lindblad, N. Sladoje. Precise Estimation of the Projection of a Shape from a Pixel Coverage Representation. In Proc. of the 7th IEEE International Symposium on Image and Signal Processing and Analysis (ISPA), Dubrovnik, Croatia. IEEE, pp. 569-574, 2011.
  • A. Tanács, J. Lindblad, N. Sladoje, and Z. Kato. Estimation of linear deformations of 3D objects. In Proc. of International Conference on Image Processing (ICIP), Hong Kong, China. IEEE, pp. 153-156, 2010.
  • N. Sladoje and J. Lindblad. High Precision Boundary Length Estimation by Utilizing Gray-Level Information. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 31, No. 2, pp. 357-363, 2009.
  • N. Sladoje, J. Lindblad. Pixel coverage segmentation for improved feature estimation. In Proc. of the 15th International Conference on Image Analysis and Processing (ICIAP),Vietri sul Mare, Italy. Lecture Notes in Computer Science, Vol. 5716, pp. 929-938, 2009.
  • A. Tanács, C. Domokos, N. Sladoje, J. Lindblad, and Z. Kato. Recovering affine deformations of fuzzy shapes. In Proc. of the 16th Scandinavian Conference on Image Analysis (SCIA), Oslo, Norway. Lecture Notes in Computer Science, Vol, 5575, pp. 735-744, 2009.


Precise image processing for biomedical and medical applications

Efficient utilization of available image data to achieve sub-pixel/voxel precision and characterize barely resolved structures, or address partial volume effect, is crucial in many biomedical and medical applications. We rely on theoretical work within the framework of discrete mathematics, which provides methods which enable preservation and efficient usage of information, aggregate information of different types, improve robustness of the developed methods and increase precision of the analysis results.

People: Joakim Lindblad, Ewert Bengtsson, Ida-Maria Sintorn, Zoltan Kato, Attila Tanács, Amit Suveer, Anindya Gupta, Patrik Malm, Anca Dragomir, Andrew Mehnert, Ramin Moshavegh

Published results:

  • A. Suveer, N. Sladoje, J. Lindblad, A. Dragomir, and I.-M. Sintorn. Enhancement of Cilia Sub-structures by Multiple Instance Registration and Super-resolution Reconstruction. In Proc. of the 20th Scandinavian Conference on Image Analysis, SCIA2017, Tromsø, Norway, Part II, Lecture Notes in Computer Science LNCS-10270, pp. 362-374, 2017.
  • A. Gupta, A. Suveer, J. Lindblad, A. Dragomir, I.-M. Sintorn, and N. Sladoje. Convolutional Neural Networks for False Positive Reduction of Automatically Detected Cilia in Low Magniffication TEM Images. In Proc. of the 20th Scandinavian Conference on Image Analysis, SCIA2017, Tromsø, Norway, Part I, Lecture Notes in Computer Science LNCS-10269, pp. 407-418, 2017.
  • A. Suveer, N. Sladoje, J. Lindblad, A. Dragomir, and I.-M. Sintorn. Automated Detection of Cilia in Low Magnification Transmission Electron Microscopy Images Using Template Matching. In Proc. of 2016 IEEE Intern. Symposium on Biomedical Imaging, ISBI2016, Prague, Czech Republic, pp. 386-390, 2016.
  • A. Tanács, J. Lindblad, N. Sladoje, and Z. Kato. Estimation of Linear Deformations of 2D and 3D Fuzzy Objects. Pattern Recognition, Vol. 48(4), pp.1387–1399, 2015.
  • J. Lindblad, E. Bengtsson, and N. Sladoje. Microscopy Image Enhancement for Cost-Effective Cervical Cancer Screening. In Proc. of the 19th Scandinavian Conference on Image Analysis (SCIA), Copenhagen, Denmark. Lecture Notes in Computer Science, Vol. 9127, pp. 440-451, 2015.
  • J. Lindblad, N. Sladoje, P. Malm, E. Bengtsson, R. Moshavegh, and A. Mehnert. Optimizing optics and imaging for pattern recognition based screening tasks. In Proc. of the 22nd Int. Conf. on Pattern Recognition (ICPR), Stockholm, Sweden, pp. 3333-3338, 2014.
  • J. Lindblad, N. Sladoje, V. Ćurić, H. Sarve, C.B. Johansson, and G. Borgefors. Improved quantification of bone remodelling by utilizing fuzzy based segmentation. In Proc. of the 16th Scandinavian Conference on Image Analysis (SCIA),Oslo, Norway. Lecture Notes in Computer Science, Vol. 5575, pp. 750-759, 2009.



Fuzzy spatial sets in image analysis

Fuzzy set theory has found many applications in image processing. I am interested in modeling image objects as spatial fuzzy sets, and developing methods applicable to such representations. This approach shows to provide information preservation and increased precision of feature estimates.

People: Joakim Lindblad, Ingela Nyström, Punam K. Saha, Marija Delić

Published results:

  • M Delić, J. Lindblad, and N. Sladoje. αLBP - a Novel Member of the Local Binary Pattern Family Based on α-cutting. In Proc. of the 9th IEEE Intern. Symposium on Image and Signal Processing and Analysis (ISPA), IEEE, Zagreb, Croatia, pp. 15-20, 2015.
  • J. Lindblad, N. Sladoje, A. Suveer, A. Dragomir, and I-M. Sintorn. High-resolution reconstruction by feature distance minimization from multiple views of an object. In Proc. of the 5th International Conference on Image Processing Theory, Tools and Applications, IPTA 2015, Orleans, France, pp. 29-34, 2015.
  • F. Malmberg, J. Lindblad, N. Sladoje, and I. Nyström. A Graph-based Framework for Sub-pixel Image Segmentation. Theoretical Computer Science, Vol. 412, No. 15, pp. 1338-1349, 2011.
  • N. Sladoje, J. Lindblad, and I. Nyström. Defuzzification of spatial fuzzy sets by feature distance minimization. Image and Vision Computing, Vol. 29, No. 2-3, pp. 127-141, 2011.
  • A. Tanács, C. Domokos, N. Sladoje, J. Lindblad, and Z. Kato. Recovering affine deformations of fuzzy shapes. In Proc. of the 16th Scandinavian Conference on Image Analysis (SCIA), Oslo, Norway. Lecture Notes in Computer Science, Vol, 5575, pp. 735-744, 2009.
  • T. Lukić, N. Sladoje, J. Lindblad. Deterministic Defuzzification based on Spectral Projected Gradient Optimization. In Proc. of 30th Symposium of the German Association for Pattern Recognition (DAGM), Munich, Germany. Lecture Notes in Computer Science, Vol. 5096, pp. 476-485, 2008.
  • N. Sladoje and J. Lindblad. Representation and Reconstruction of Fuzzy Disks by Moments. Fuzzy Sets and Systems, Vol. 158, No. 5, pp. 517-534, 2007.
  • J. Lindblad, T. Lukić, and N. Sladoje. Defuzzification by Feature Distance Minimization Based on DC Programming. In Proc. of 5th International Symposium on Image and Signal Processing and Analysis (ISPA),Istanbul, Turkey. IEEE, pp. 373-378, 2007.
  • J. Lindblad, N. Sladoje, and T. Lukić. Feature Based Defuzzification in Z2 and Z3 Using a Scale Space Approach. In Proc. of 13th International Conference on Discrete Geometry for Computer Imagery (DGCI 2006), Szeged, Hungary. Lecture Notes in Computer Science, Vol. 4245, pp. 379-390, 2006.
  • J. Lindblad and N. Sladoje. Feature Based Defuzzification at Increased Spatial Resolution. In Proc. of 11th International Workshop on Combinatorial Image Analysis (IWCIA2006),Berlin,Germany. Lecture Notes in Computer Science, Vol. 4040, pp.131-143, 2006.
  • J. Chanussot, I. Nyström and N. Sladoje. Shape Signatures of Fuzzy Star-shaped Sets Based on Distance from the Centroid. Pattern Recognition Letters, Vol. 26(6), pp. 735-746, 2005.
  • N. Sladoje, I. Nyström, and P.K. Saha. Measurements of digitized objects with fuzzy borders in 2D and 3D. Image and Vision Computing, Vol. 23, pp 123-132, 2005.
  • N. Sladoje and J. Lindblad. Estimation of Moments of Digitized Objects with Fuzzy Borders. In Proc. of International Conference on Image Analysis and Processing (ICIAP2005), Cagliari, Italy. Lecture Notes in Computer Science, Vol. 3617, pp. 188-195, 2005.
  • N. Sladoje, J. Lindblad and I. Nyström. Defuzzification of Discrete Objects by Optimizing Area and Perimeter Similarity. In Proc. of 17th International Conference on Pattern Recognition (ICPR 2004), Cambridge, UK. IEEE, Vol. 3, pp 526-529, 2004.
  • N. Sladoje, I. Nyström, and P.K. Saha. Perimeter and area estimations of digitized objects with fuzzy border. In Proc. of Discrete Geometry for Computer Imagery (DGCI 2003), Naples, Italy. Lecture Notes in Computer Science, Vol. 2886, pp 368-377, 2003.
  • N. Sladoje, I. Nyström, and P.K. Saha. Measuring perimeter and area in low resolution images using a fuzzy approach. In Proc. of 13th Scandinavian Conference on Image Analysis (SCIA 2003), Göteborg, Sweden. Lecture Notes in Computer Science, Vol. 2749, pp 853-860, 2003.


Shape representation and reconstruction

My early interests were representation and reconstruction of binary shapes by geometric moments.

People: Joviša Žunić

Published results:

  • J. Zunic and N. Sladoje. Efficiency of Characterizing Ellipses and Ellipsoids by Discrete Moments. IEEE Trans. Pattern Analysis and Machine Intelligence, Vol.22, No.4, pp. 407-414, 2000.
  • N. Sladoje. The Reconstruction of the Digital Hyperbola Segment from its Code. In Proc. of Discrete Geometry for Computer Imagery (DGCI 2000), Uppsala, Sweden. Lecture Notes in Computer Science, Vol. 1953, pp. 159-170, 2000.
  • N. Sladoje and J. Zunic. A reconstruction of digital parabolas from their least squares fit representation. Yugoslav Journal of Operations Research, Vol.7 No.2, pp 1-14, 1997.
  • N. Sladoje and J. Zunic. Ellipses Estimation from their Digitization. In Proc. of Discrete Geometry for Computer Imagery (DGCI 1997), Montpellier, France. Lecture Notes in Computer Science, Vol.1347, pp. 187-198, 1997.
  • J. Zunic and N. Sladoje. A Characterization of Digital Disks by Discrete Moments. In Proc. of Computer Analysis of Images and Patterns (CAIP1997), Kiel, Germany. Lecture Notes in Computer Science, Vol.1296, pp. 582-589, 1997.



Nataša Home