Refereed conference proceedings

1. Distance Measures Between Digital Fuzzy Objects and their Applicability in Image Processing
   Authors: Vladimir Curic, Joakim Lindblad, and Natasa Sladoje (1)
   (1) Faculty of Technical Sciences, University of Novi Sad, Novi Sad, Serbia
   Conference: 14th International Workshop on Combinatorial Image Analysis (IWCIA), Madrid, Spain, volume 6636 of Lecture Notes in Computer Science, pp 385-397
   Publisher: Springer, Berlin/Heidelberg
   Editors: Jake Aggarwal, Reneta Barneva, Valentin Brimkov, Kostadin Koroutchev, and Elka Koroutcheva
   Abstract: We present two different extensions of the Sum of minimal distances and the Complement weighted sum of minimal distances to distances between fuzzy sets. We evaluate to what extent the proposed distances show monotonic behavior with respect to increasing translation and rotation of digital objects, in noise free, as well as in noisy conditions. Tests show that one of the extension approaches leads to distances exhibiting very good performance. Furthermore, we evaluate distance based classification of crisp and fuzzy representations of objects at a range of resolutions. We conclude that the proposed distances are able to utilize the additional information available in a fuzzy representation, thereby leading to improved performance of related image processing tasks.

2. Precise Estimation of the Projection of a Shape from a Pixel Coverage Representation
   Authors: Slobodan Drazic (1), Joakim Lindblad, and Natasa Sladoje (1)
   (1) Faculty of Technical Sciences, University of Novi Sad, Serbia
   Conference: 7th IEEE International Symposium on Image and Signal Processing and Analysis (ISPA), Dubrovnik, Croatia, pp 569-574
   Publisher: IEEE Computer Society, Piscataway, NJ
   Abstract: Measuring width and diameter of a shape are problems well studied in the literature. A pixel coverage representation is one specific type of digital fuzzy representation of a continuous image object, where the (membership) value of each pixel is (approximately) equal to the relative area of the pixel which is covered by the continuous object. Lately a number of methods for shape analysis use pixel coverage for reducing error of estimation. We introduce a novel method for estimating the projection of a shape in a given direction. The method is based on utilizing pixel coverage representation of a shape. Performance of the method is evaluated by a number of tests on synthetic objects, confirming high precision and applicability for calculation of diameter and elongation of a shape.

3. Horizontal Features Based Illumination Normalization Method for Face Recognition
   Authors: Muhammad Talal Ibrahim (1), M. Khalid Khan Niazi, and Ling Guan (1)
   (1) Ryerson Multimedia Lab, Toronto, Canada
   Conference: 7th IEEE International Symposium on Image and Signal Processing and Analysis (ISPA), Dubrovnik, Croatia, pp 684-689
   Publisher: IEEE Computer Society, Piscataway, NJ
   Abstract: This paper presents a novel filtering method for face recognition under varying illumination. The proposed method starts by normalizing the given input image by gamma transformation. The shadow artifacts in the normalized image are reduced with the decimation free directional filter banks (DDFB). We have used correlation coefficient as a similarity measure for face recognition. Empirically, we have proven that most of the discriminating features in a human face are horizontal in nature. The efficiency of the proposed method is evaluated on two public databases: Yale Face Database B, and the Extended Yale Face Database B. Experimental results demonstrate that the proposed method achieves higher recognition rate under varying illumination conditions in comparison with some other existing methods.

4. Virus Texture Analysis Using Local Binary Patterns and Radial Density Profiles
   Authors: Gustaf Kylberg, Mats Uppström (1), and Ida-Maria Sintorn
   (1) Vironova AB, Stockholm
   Conference: 16th Iberoamerican Congress on Pattern Recognition (CIARP), Pucón, Chile, volume 7042 of Lecture Notes in Computer Science, pp 573-580
   Publisher: Springer, Berlin/Heidelberg
   Editors: César San Martin and Sang-Woon Kim
   Abstract: We investigate the discriminant power of two local and two global texture measures on virus images. The viruses are imaged using negative stain transmission electron microscopy. Local binary patterns and a multi scale extension are compared to radial density profiles in the spatial domain and in the Fourier domain. To assess the discriminant potential of the texture measures a Random Forest classifier is used. Our analysis shows that the multi scale extension performs better than the standard local binary patterns and that radial density profiles in comparison is a rather poor virus texture discriminating measure. Furthermore, we show that the multi scale extension and the profiles in Fourier domain are both good texture measures and that they complement each other well, that is, they seem to detect different texture properties. Combining the two, hence, improves the discrimination between virus textures.

5. Generalized Hard Constraints for Graph Segmentation
   Authors: Filip Malmberg, Robin Strand, and Ingela Nyström
   Conference: 17th Scandinavian Conference on Image Analysis (SCIA), Ystad Saltsjöbad, volume 6688 of Lecture Notes in Computer Science, pp 36-47
   Publisher: Springer, Berlin/Heidelberg
   Editors: Anders Heyden and Fredrik Kahl
   Abstract: Graph-based methods have become well-established tools for image segmentation. Viewing the image as a weighted graph, these methods seek to extract a graph cut that best matches the image content. Many of these methods are interactive, in that they allow a human operator to guide the segmentation process by specifying a set of hard constraints that the cut must satisfy. Typically, these constraints are given in one of two forms: regional constraints (a set of vertices that must be separated by the cut) or boundary constraints (a set of edges that must be included in the cut). Here, we propose a new type of hard constraints, that includes both regional constraints and boundary constraints as special cases. We also present an efficient method for computing cuts that satisfy a set of generalized constraints, while globally minimizing a graph cut measure.

6. Image Foresting Transform: On-the-fly Computation of Segmentation Boundaries
   Author: Filip Malmberg
   Conference: 17th Scandinavian Conference on Image Analysis (SCIA), Ystad Saltsjöbad, volume 6688 of Lecture Notes in Computer Science, pp 616-624
   Publisher: Springer, Berlin/Heidelberg
   Editors: Anders Heyden and Fredrik Kahl
   Abstract: The Image Foresting Transform (IFT) is a framework for seeded image segmentation, based on the computation of minimal cost paths in a discrete representation of an image. In two recent publications, we have shown that the segmentations obtained by the IFT may be improved by refining the segmentation locally around the boundaries between segmented regions. Since these methods operate on a small sub-set of the image elements only, they may be implemented efficiently if the set of boundary elements is known. Here, we show that this set may be obtained on-the-fly, at virtually no additional cost, as a by-product of the IFT algorithm.

7. Bias Field Correction Using Grey-Weighted Distance Transform Applied on MR Volumes
   Authors: M. Khalid Khan Niazi, Ingela Nyström, Muhammad Talal Ibrahim (1), and Ling Guan (1)
   (1) Ryerson Multimedia Lab, Toronto, Canada
   Conference: 8th International Symposium on Biomedical Imaging (ISBI), Barcelona, Spain, pp 357-360
   Publisher: IEEE Computer Society, Piscataway, NJ
   Abstract: We present a method that produces data-dependent high-pass filters that can be used in, e.g., image restoration. MR volumes often suffer from bias field artifacts produced due to unsteady magnetic field and spatial variations in reception of the RF magnetic eld emitted by the subject. To cope with these artifacts, we propose a method based on the grey-weighted distance transform (GWDT). It first computes the spectrum of the MR volume using Fast Fourier Transform, and then computes the GWDT of the magnitude spectrum. The power spectrum computed using GWDT helps in designing a high-pass filter which is later used to correct the bias artifacts. Experimental results on close to 100 datasets demonstrate the effectiveness of our method.

8. Robust Signal Generation and Analysis of Rat Embryonic Heart Rate In Vitro using Laplacian Eigenmaps and Empirical Mode Decomposition
   Authors: M. Khalid Khan Niazi, Muhammad Talal Ibrahim (1), Mats F. Nilsson, Anna-Carin Sköld, Ling Guan (1), and Ingela Nyström
   (1) Ryerson Multimedia Lab, Toronto, Canada
   Conference: 14th International Conference on Computer Analysis of Images and Patterns (CAIP), Seville, Spain, volume 6855 of Lecture Notes in Computer Science, pp 523-530
   Publisher: Springer, Berlin/Heidelberg
   Editors: Pedro Real, Daniel Diaz-Pernil, Helena Molina-Abril, Ainhoa Berciano, and Walter Kropatsch
   Abstract: To develop an accurate and suitable method for measuring the embryonic heart rate in vitro, a system combining Laplacian eigenmaps and empirical mode decomposition has been proposed. The proposed method assess the heart activity in two steps; signal generation and heart signal analysis. Signal generation is achieved by Laplacian eigenmaps (LEM) in conjunction with correlation co-efficient, while the signal analysis of the heart motion has been performed by the modified empirical mode decomposition (EMD). LEM helps to find the template for the atrium and the ventricle respectively, whereas EMD helps to find the non-linear trend term without defining any regression model. The proposed method also removes the motion artifacts produced due to the the non-rigid deformation in the shape of the embryo, the noise induced during the data acquisition, and the higher order harmonics. To check the authenticity of the proposed method, 151 videos have been investigated. Experimental results demonstrate the superiority of the proposed method in comparison to three recent methods.

9. Path-Based Distance with Varying Weights and Neighborhood Sequences
   Authors: Nicolas Normand (1,2), Robin Strand, Pierre Evenou (1), and Aurore Arlicot (1)
   (1) IRCCyN, University of Nantes, France
   (2) School of Physics, Monash University, Melbourne, Australia
   Conference: 16th International Conference on Discrete Geometry for Computer Imagery (DGCI), Nancy, France, volume 6607 of Lecture Notes in Computer Science, pp 199-210
   Publisher: Springer, Berlin/Heidelberg
   Editors: Isabelle Debled-Rennesson, Eric Domenjoud, Bertraud Kerautret, and Philippe Even
   Abstract: This paper presents a path-based distance where local displacement costs vary both according to the displacement vector and with the travelled distance. The corresponding distance transform algorithm is similar in its form to classical propagation-based algorithms, but the more variable distance increments are either stored in look-up-tables or computed on-the-fly. These distances and distance transform extend neighborhood-sequence distances, chamfer distances and generalized distances based on Minkowski sums. We introduce algorithms to compute, in , a translated version of a neighborhood sequence distance map with a limited number of neighbors, both for periodic and aperiodic sequences. A method to recover the centered distance map from the translated one is also introduced. Overall, the distance transform can be computed with minimal delay, without the need to wait for the whole input image before beginning to provide the result image.

10. Visualization and Haptics for Interactive Medical Image Analysis: Image Segmentation in Cranio-Maxillofacial Surgery Planning
    Authors: Ingela Nyström, Johan Nysjö, and Filip Malmberg
    Conference: 2nd International Visual Informatics Conference (IVIC), Visual Informatics: Sustaining Research and Innovations, Selangor, Malaysia, volume 7066 of Lecture Notes in Computer Science, pp 1-12
    Publisher: Springer, Berlin/Heidelberg
    Editors: Halimah Badioze Zaman, Peter Robinson, Maria Petrou, Patrick Olivier, Timothy K. Shih, Sergio Velastin, and Ingela Nyström
    Abstract: A central problem in cranio-maxillofacial (CMF) surgery is to restore the normal anatomy of the skeleton after defects, e.g., trauma to the face. With careful pre-operative planning, the precision and predictability of the craniofacial reconstruction can be significantly improved. In addition, morbidity can be reduced thanks to shorter operation time. An important component in surgery planning is to be able to accurately measure the extent of anatomical structures. Of particular interest are the shape and volume of the orbits (eye sockets). These properties can be measured in 3D CT images of the skull, provided that an accurate segmentation of the orbits is available. Here, we present a system for interactive segmentation of the orbit in CT images. The system utilizes 3D visualization and haptic feedback to facilitate efficient exploration and manipulation of 3D data.
11. High-Throughput Cellular-Resolution in vivo Vertebrate Screening
    Authors: Carlos Pardo-Martin (1,2), Tsung-Yao Chang (1), Amin Allalou, Carolina Wählby, and Mehmet Fatih Yanik (1,3)
    (1) Massachusetts Institute of Technology (MIT), MA, USA
    (2) Harvard University, Moston, MA, USA
    (3) Broad Institute, Boston, MA, USA
    Conference: 15th International Conference on Miniaturized Systems for Chemistry and Life Sciences (TAS), Seattle, WA, USA, pp 1557-1559
    Publisher: Chemical and Biological Microsystems Society (CBMS)
    Abstract: Due to its small size and optical transparency, zebrafish larvae are excellent vertebrate models to study human diseases in vivo. We previously developed a high throughput screening platform capable of handling, cellular-resolution imaging and optically manipulating zebrafish larvae. Here, we present new novel technologies to significantly increase the throughput of our screening platform by multi-threading the loading and imaging processes and high-speed algorithms to automatically manipulate the larvae.

12. Quantification of Gaseous Structures with Volumetric Reconstruction from Visual Hulls
    Authors: Stefan Seipel and Peter Jenke (1)
    (1) Faculty of Engineering and Sustainable Technology, University of Gävle
    Conference: SIGRAD, Swedish Chapter of Eurographics, Stockholm, Sweden, pp 77-82
    Publisher: Linköping University Electronic Press, Linköping University
    Abstract: 3D reconstruction from visual hulls is a well established technique for camera based reconstruction of 3D objects in computer graphics. We propose in this paper to employ visual hull techniques to quantify the volume of diffusely defined gaseous structures. In our evaluation, visual quality of the 3D reconstructions is secondary. Instead, using synthetic ground truth data, we determine the number of independent silhouette images needed to achieve a stable volume estimate. We also estimate the influence of different segmentation results of the silhouette images on final volume estimates. Our results show that comparably few camera views yield to convergent volume estimates. For the type of 3D data studied, visual hull reconstructions overestimate actual volumes with about 50%. This proportion seems to be consistent for different data sets tested and may serve for re-calibration of volume estimation of gaseous structures.

13. A Weighted Neighbourhood Sequence Distance Function with Three Local Steps
    Authors: Robin Strand and Benedek Nagy (1)
    (1) Dept. of Computer Science, Faculty of Informatics, University of Debrecen, Hungary
    Conference: 7th IEEE International Symposium on Image and Signal Processing and Analysis (ISPA), Dubrovnik, Croatia, pp 564-568
    Publisher: IEEE Computer Society, Piscataway, NJ
    Abstract: We present a combined weighted neighborhood sequence distance function on the square grid with three types of steps. For this general distance function, we compute parameters that optimize an error function for the asymptotic shape of digital disks. We also analyze approximations of the parameters that can be used in the digital grid used here. An algorithm that can be used for image processing applications is also presented.

14. Sparse Object Representations by Digital Distance Functions
    Author: Robin Strand
    Conference: 16th International Conference on Discrete Geometry for Computer Imagery (DGCI), Nancy, France, volume 6607 of Lecture Notes in Computer Science, pp 211-222
    Publisher: Springer, Berlin/Heidelberg
    Editors: Isabelle Debled-Rennesson, Eric Domenjoud, Bertraud Kerautret, and Philippe Even
    Abstract: In this paper, some methods for representing objects using path-based distances are considered. The representations can be used as anchor points when extracting medial representations of the objects. The distance transform (DT) is obtained by labeling each object element with the distance to the background. By local operations on the DT, different sets of anchor points can be obtained. We present two different methods based on local operations and prove that the representations are reversible, when this is the case. The methods are defined for weighted distances based on neighborhood sequences, which includes for example the well known cityblock and chessboard distances.
15. Investigating Measures for Transfer Function Generation for Visualization of MET Biomedical Data
    Authors: Lennart Svensson, Ingela Nyström, Stina Svensson, and Ida-Maria Sintorn
    Conference: 19th International Conference on Computer Graphics, Visualization and Computer Vision (WSCG), Plzen, Czech Republic, Communication Papers Proceedings, pp 113-120
    Publisher: Union Agency, Plzen
    Editors: Gladimir Baranoski and Vaclav Skala
    Abstract: In this paper, the question of automatically setting transfer functions for volume images is further explored. Morespecifically, the focus is automatic visualization of Molecular Electron Tomography (MET) volume images usingone-dimensional transfer functions. We investigate how well a few general measures based on density, gradient,curvature and connected component information are suited for generating these transfer functions. To assessthis, an expert has set suitable transfer function levels manually and we have studied how these levels relate todifferent characteristics of the selected measures for 29 data sets. We have found that the measures can be used toautomatically generate a transfer function used to visualize MET data, to give the user an approximate view of thecomponents in the image.

16. Registration Parameter Spaces for Molecular Electron Tomography Images
    Authors: Lennart Svensson, Anders Brun, Ingela Nyström, and Ida-Maria Sintorn
    Conference: 16th International Conference on Image Analysis and Processing (ICIAP), Ravenna, Italy, part I, volume 6978 of Lecture Notes in Computer Science, pp 403-412
    Publisher: Springer, Berlin/Heidelberg
    Editors: Giuseppe Maino and Gian Luca Foresti
    Abstract: We describe a methodology for exploring the parameter spaces of rigid-body registrations in 3-D. It serves as a tool for guiding and assisting a user in an interactive registration process. An exhaustive search is performed over all positions and rotations of the template, resulting in a 6-D volume, or fitness landscape. This is explored by the user, who selects and views suitable 3-D projections of the data, visualized using volume rendering. The 3-D projections demonstrated here are the maximum and average intensity projections of the rotation parameters and a projection of the rotation parameter for fixed translation parameters. This allows the user to jointly visualize projections of the parameter space, the local behaviour of the similarity score, and the corresponding registration of the two volumes in 3-D space for a chosen point in the parameter space. The procedure is intended to be used with haptic exploration and interaction. We demonstrate the methodology on a synthetic test case and on real molecular electron tomography data using normalized cross correlation as similarity score.

17. Data Mining Medieval Documents by Word Spotting
    Authors: Fredrik Wahlberg, Mats Dahllöf (1), Lasse Mårtensson (2), and Anders Brun
    (1) Dept. of Linguistics and Philology, UU
    (2) Dept. of Scandinavian Languages, UU
    Conference: 1st International Workshop on Historical Document Imaging and Processing (HIP), Beijing, China, pp 75-82
    Publisher: ACM, New York
    Abstract: This paper presents novel results for word spotting based on dynamic time warping applied to medieval manuscripts in Latin and Old Swedish. A target word is marked by a user, and the method automatically finds similar word forms in the document by matching them against the target. The method automatically identifies pages and lines. We show that our method improves accuracy compared to earlier proposals for this kind of handwriting. An advantage of the new method is that it performs matching within a text line without presupposing that the difficult problem of segmenting the text line into individual words has been solved. We evaluate our word spotting implementation on two medieval manuscripts representing two script types. We also show that it can be useful by helping a user find words in a manuscript and present graphs of word statistics as a function of page number.

18. Accurate Estimation of Gaussian and Mean Curvature in Volumetric Images
    Authors: Erik L. G. Wernersson, Cris L. Luengo Hendriks, and Anders Brun
    Conference: International Conference on 3D Imaging, Modeling, Processing, Visualization, and Transmission (3DIMPVT), Hangzhou, China, pp 312-317
    Publisher: IEEE Computer Society, Piscataway, NJ
    Abstract: Curvature is a useful low level surface descriptor of wood fibres in 3D micro-CT images of paper and composite materials. It may for instance be used to differentiate between the outside and the inside (lumen) of wood fibre. Since the image acquisition introduces noise, some kind of smoothing is required to obtain accurate estimates of curvature. However, in these materials, the fibres of interest are frequently both thin and densely packed. In this paper, we show how existing methods fail to accurately capture curvature information under these circumstances. Maintained resolution and smoothing of noise are two competing goals. In some situations, existing methods will even estimate the wrong signs of the principal curvatures. We also present a novel method, which is shown to have better performance in several experiments. This new method will generically produce better curvature estimates for thin objects and objects in close proximity.

19. Indication of Methane Gas in IR-Imagery
    Authors: Julia Åhlén (1) and Stefan Seipel
    (1) Dept. of Industrial Development, IT and Land Management, University of Gävle
    Conference: IADIS International Conference on Computer Graphics, Visualization, Computer Vision and Image Processing (CGVCVIP), Rome, Italy, pp 187-182 (electronic publication)
    Publisher: IADIS Press
    Editor: Yingcay Xiao
    Abstract: There are human produced sources of methane gas, such as waste storages, that contribute to the global warmth and other negative effects. There is not much research on the correlation of such leakage and greenhouse effect. Methane gas is not visible for humans and thus impossible to detect using commercial cameras. Specially designed IR-camera can detect this gas and thus is used in this study. Using digital video taken over a waste disposal place we create a detection algorithm that is sensitive to the spectral and morphological characteristics of methane gas. In case of small spot leakage there is a reason to assume failure in piping system and in case of widely spread leakage area we can state that it is caused by unsupervised storage of waste and this should be attended immediately. Background and target gas are distinguished using spectral and morphological classifiers, which are extracted from the analyzed IR-imagery. It is shown that methane gas detection can be carried out efficiently using image processing techniques and the definition of turbulence of the image.