[MGV logo]   Vol. 18 (2009):
  Abstracts of Papers

No. 1,
No. 2: Special Issue on Special Issue on Application of Novel Classification Methods in Computer Vision,
No. 3,
No. 4.

17 (2008) main 19 (2010)

Machine GRAPHICS & VISION, Vol. 18 (2009), No. 1:

Jourdan G.-V., Rakotomalala L., Zaguia N.:
LR-Upward Drawing of Ordered Sets.
MGV vol. 18, no. 1, 2009, pp. 3-19.
In this paper we introduce a new concept to draw an ordered set: the \textit{LR-Upward} drawing is an upward drawing based on a chain decomposition of the order such that elements drawn on the same vertical line are always comparable and all other comparabilities flow from left to right. We describe a particular technique for automatic generation of an enhanced \textit{LR-Upward} drawing for \textit{N-Free} orders that are \textit{X-Cycle-Free}. This technique first enhances the drawing locally, around a particular chain, and then expands the enhancement to the remaining part of the order. A Java-based implementation is also presented.
Key words: Ordered sets, Ordered sets visualization, Upward drawings, N-Free, X-Cycle-Free.

Beszédes M., Culverhouse P., Oravec M.:
Facial Emotion Classification Using Active Appearance Model and Support Vector Machine Classifier.
MGV vol. 18, no. 1, 2009, pp. 21-46.
Automatic analysis of human face expression is an interesting and non-trivial problem. In the last decade, many approaches have been described for emotion recognition based on analysis of facial expression. However, little has been done in the sub-area of the recognition of facial emotion intensity levels. This paper proposes the analysis of the use of Active Appearance Models (AAMs) and Support Vector Machine (SVM) classifiers in the recognition of human facial emotion and emotion intensity levels. AAMs are known as a tool for statistical modeling of object shape/appearance or for precise object feature detection. In our case, we examine their properties as a technique for feature extraction. We analyze the influence of various facial feature data types (shape / texture / combined AAM parameter vectors) and the size of facial images on the final classification accuracy. Then, approaches to proper C-SVM classifiers (RBF kernel) training parameter adjustment are described. Moreover, an alternative way of classification accuracy evaluation using the human visual system as a reference point is discussed. Unlike the usual to the approach evaluation of recognition algorithms (based on comparison of final classification accuracies), the proposed evaluation schema is independent of the testing set parameters, such as number, age and gender of subjects or the intensity of their emotions. Finally, we show that our automatic system gives emotion categories for images more consistent labels than human subjects, while humans are more consistent in identifying emotion intensity level compared to our system.
Key words: automatic facial emotion recognition, active appearance model, support vector machine, psychological experiment, cross-validation.

Shahbe M.D., Hati S.:
Kernel Based Subspace Methods: Infrared vs Visible Face Recognition.
MGV vol. 18, no. 1, 2009, pp. 47-66.
This paper investigates the use of kernel theory in two well-known, linear-based subspace representations: Principle Component Analysis (PCA) and Fisher's Linear Discriminant Analysis (FLD). The kernel-based method provides subspaces of high-dimensional feature spaces induced by some nonlinear mappings. The focus of this work is to evaluate the performances of Kernel Principle Component Analysis (KPCA) and Kernel Fisher's Linear Discriminant Analysis (KFLD) for infrared (IR) and visible face recognition. The performance of the kernel-based subspace methods is compared with that of the conventional linear algorithms: PCA and FLD. The main contribution of this paper is the evaluation of the sensitivities of both IR and visible face images to illumination conditions, facial expressions and facial occlusions caused by eyeglasses using the kernel-based subspace methods.
Key words: Face Recognition; Principle Component Analysis; Linear Discriminant Analysis, Kernel Methods.

Holinka S., Sára R., Smutek D.:
Relation Between Structural Changes in B-mode Ultrasound Images of Thyroid Parenchyma and the Presence of Thyroid Antibodies in Blood Sample.
MGV vol. 18, no. 1, 2009, pp. 67-82.
Ultrasonography is a cheap and quick non-invasive medical imaging technique, used as a diagnostic method for autoimmune thyroiditis. Another important diagnostic method for this chronic inflammation is measuring the increased level of thyroid antibodies (anti-thyroid peroxidase (TPOAb) and/or anti- thyroglobulin (TgAb)) in blood samples. This paper shows that B-mode ultrasound images contain weak information related to the presence or absence of these antibodies. Ultrasound image analysis is based on textural recognition using probabilistic spatial features. Two studies are performed. The results of the first study show that the spatial texture features we used contain weak information about the presence or absence of TPOAb and TgAb antibodies as measured by conditional entropy. In the second study, a classifier derived from Bayesian decision theory is tested on a set of 2820 sonograms of 94 subjects. The training set contains 67 subjects and the test set consists of 27 independent subjects. The results of classification to three classes (healthy, thyroiditis with positive antibody test, thyroiditis with negative antibody test) achieved sensitivity 29% and specificity 100% on the test set.
Key words: B-mode sonography, texture analysis, classification, thyroid gland, autoimmune lymphocytic thyroiditis, thyroid antibodies.

Aouat S., Larabi S.:
Comparison of Detailed Descriptors for Noisy Silhouettes.
MGV vol. 18, no. 1, 2009, pp. 83-104.
In this paper we propose a new method for comparing silhouettes. Silhouettes of 3D objects, extracted from 2D images, are described using the LWDOS language \cite {1}. Sometimes, even though the silhouettes appear identical, their LWDOS detailed descriptors may be very different; therefore, we might wrongly conclude that the two silhouettes are different. In order to eliminate this problem which is due to several phenomena, we try to deduce the same detailed descriptor from two slightly different descriptors of two silhouettes.
Key words: Detailed Descriptor, Silhouette, LWDOS, Comparison.

Sharma A., Sharma R.K., Kumar R.:
Online Preprocessing of Handwritten Gurmukhi Strokes.
MGV vol. 18, no. 1, 2009, pp. 105-120.
In this paper, the authors have implemented preprocessing algorithms for online handwritten Gurmukhi strokes in order to find the improvements in recognition of four high-level features (loop, headline, straight line and dot) of Gurmukhi strokes. Preprocessing algorithms include size normalization and centering, interpolating missing points, smoothing, slant correction and resampling of points. Recognition algorithms for the above mentioned four high-level features are also introduced in this paper. Experiments have been conducted across 60 writers and 5%, 3.33%, 6.66% and 8.34% improvements have been observed for recognition of loop, headline, straight line and dot features, respectively, after using preprocessing algorithms.
Key words: Recognition, Preprocessing, High-Level features.

Machine GRAPHICS & VISION, Vol. 18 (2009), No. 2:

Special Issue on Application of Novel Classification Methods in Computer Vision
Special Issue Editor: Katarzyna Stapor.

Koczkodaj W.W., Robidoux N., Tadeusiewicz R.:
Classifying visual objects with the consistency-driven pairwise comparisons method.
MGV vol. 18, no. 2, 2009, pp. 143-153.
The classification of the various image features or visual objects can be carried out by the consistency-driven pairwise comparisons method based on their relative importance. A key issue in the proposed approach is a weight-based synthesis for combining various image features. When compared with the traditional experience-based linear assignment method, the proposed approach is more effective and easy to communicate.
Key words: machine vision, visual object, pairwise comparisons, inconsistency analysis, content-based image retrieval.

Iwanowski M.:
Morphological Classification of Binary Image's Pixels.
MGV vol. 18, no. 2, 2009, pp. 155-174.
The paper presents a novel approach to the classification of binary image pixels based on mathematical morphology. The proposed method makes use of new {\em class extractor} operators to extract pixels belonging to pre-defined classes from the original image. The classes consist of pixels characterized by particular morphological properties. In the paper, a general scheme of a morphological classifier is introduced. Depending on the class to be detected, various extractors can be defined based on the mathematical morphology. The extracted classes can be organized in a class hierarchy tree structure. Apart from the general framework of a morphological classifier, various class extractors are described a well. Another notion introduced in the paper are class distribution functions, which can be used as shape descriptors for pattern recognition. Finally, two practical examples of classification of binary image pixels are presented, representing an application to shape description and to analysis of the shape of water bodies.
Key words: mathematical morphology, image classification, binary images.

Gridin V.N., Titov V.S., Truphanov M.I., Korostelev S.I.:
Vision System for Image Recognition Based on Three-dimensional Vector Patterns.
MGV vol. 18, no. 2, 2009, pp. 175-.
We introduce a vision system for object image recognition based on a three-dimensional vector pattern, which offers a possibility to solve a broad variety of problems connected with image recognition. We point out that the use of three-dimensional vector models as patterns enables substantial reduction in the memory costs and simplification of the learning process. In the paper we provide a description of the image recognition method implemented by the vision system, and propose the ways for its development.
Key words: Vision System, Image Recognition, Three-dimensional Vector Patterns, 3D pattern.

Amine A., Ghouzali S., Rziza M., Aboutajdine D.:
An improved Method for Face Recognition Based on SVM in Frequency Domain.
MGV vol. 18, no. 2, 2009, pp. 187-.
We examine the problem of discriminating between objects of more than two classes using 'minimum information'. Discrete Cosine Transforms (DCT) represents a computationally simple and efficient method that preserves the structure of the data without introducing significant distortion. In this paper, an efficient face recognition method combined DCT and Support Vector Machine (SVM) is proposed. The underlying algorithm is derived by applying DCT to several regions of a face image. Only a small subset of the DCT coefficients is retained by truncating high frequency DCT components in each block. Selected DCT features are then subjected to SVM for class separability enhancement before being used for face recognition. This leads to a new, low-dimensional representation of images which allows for a fast and simple classification. In this context, we have performed a large number of experiments using two popular face databases: ORL and Yale, and comparisons using PCA, LDA, ICA, MLP, etc. Experimental results show that the proposed method performs better than traditional approaches in terms of both efficiency and accuracy.
Key words: Face recognition, Feature Extraction, Support Vector Machine, Discrete Cosine Transform.

Ghouzali S., Hemami S., Rziza M., Aboutajdine D.:
Generalized Gaussian Density for Skin Detection in DCT Domain.
MGV vol. 18, no. 2, 2009, pp. 201-.
In this paper, we propose a highly efficient algorithm to model the human skin color. The algorithm involves generating a discrete Cosine transform (DCT) at each pixel location, using the surrounding points. The DCT coefficients incorporate the pixel color and texture information to distinguish between skin and non-skin. A generalized Gaussian distribution (GGD) is used in this framework to model the DCT coefficients at low frequencies. Next, the model parameters are estimated using the maximum-likelihood (ML) criterion applied to a set of training skin samples. Finally, each pixel is classified as skin if its likelihood ratio exceeds some threshold. The experimental results show that our model avoids excessive false detection while still retaining a high degree of correct detection.
Key words: Skin detection, Generalized Gaussian distribution, Discrete Cosine Transform, Image segmentation.

Mahersia H., Hamrouni K.:
Rotation- and Scale-Invariant Texture Classification Using Log-Polar and Ridgelet Transforms.
MGV vol. 18, no. 2, 2009, pp. 215-.
Classification of distorted texture images is a challenging and important problem in real world image analysis and understanding. This paper proposes a new texture characterization method which is robust to geometric distortions, including rotation and scale changes. The rotation- and scale-invariant feature extraction for a given image involves applying the log-polar transform to eliminate the rotation and scale effects, followed by the ridgelet transform. In the experiments, the K-nearest neighborhood classifier is employed, using Euclidian and Manhattan distances to classify two sets of 30 and 40 distinct natural textures selected from the Brodatz and the VisTex albums. The experimental results, based on different test data sets for images with different orientations and scales, show that the proposed classification scheme using log-polar ridgelet signatures outperforms texture classification based on log-polar and wavelet transforms. Its overall accuracy rate reaches 100% for orientation or scale changes, and is about 73.708\% for joint rotation and scale changes. These results demonstrate the effectiveness of our characterization method in texture image classification experiments.
Key words: Classification, Log-polar transform, Segmentation, Ridgelet transform, Texture.

Tominaga S.:
Material Classification Method for Printed Circuit Boards Using a Spectral Imaging System.
MGV vol. 18, no. 2, 2009, pp. 233-.
This paper proposes a method for classifying object materials on a raw circuit board into element materials by means of surface-spectral reflectance. First, we develop a spectral imaging system for observing the minute details of the board and capturing their spectral data. Second, the surface-spectral reflectance functions of the board are estimated by a direct method using narrow band sensor outputs. We investigate the reflection properties of various objects on the board under different illumination directions. Third, we find key features of the body spectral reflectances for different materials, and present a rule for classifying the objects into six element materials. Finally, experiments are executed using a real circuit board. The observed spectral reflectance image is segmented into the element material areas. The performance and robustness of the proposed method are examined in detail in comparison with other methods.
Key words: spectral imaging, material classification, raw circuit boards, surface-spectral reflectance.

Machine GRAPHICS & VISION, Vol. 18 (2009), No. 3:

Zelawski M.:
Detecting Pathologies with Homology Algorithms in Magnetic Resonance Images of Brain.
MGV vol. 18, no. 3, 2009, pp. 253-266.
This paper describes a method for detecting the presence of pathological changes in two-dimensional brain images from magnetic resonance examination. The proposed idea is based on homology theory, which makes it easily extendable to three-dimensional brain images in particular it may be applied to the computer tomography data.
Key words: homology algorithm, brain tumor, MR imaging, medical imaging.

Abdel-Qader I., Zyout I., Jacobs C.:
Detection of Synthetic and Real Microcalcifications based on Statistical Analysis of Original and Highpass-Filtered Mammograms.
MGV vol. 18, no. 3, 2009, pp. 267-288.
Detection of clustered microcalcifications in digitized mammograms can be very useful for early detection of breast cancer. Clustered microcalcifications have a distinguished signature in both spatial and frequency domains. In the spatial domain, they appear as white spots which represent local maxima, while in the frequency domain microcalcifications represent local anomalies that can be captured within the high frequency subbands. In this work, we propose an algorithm for detection of clustered microcalcifications by utilizing these signatures, integrating the statistical parameters of both spatial and frequency domains. The results prove the effectiveness of the proposed method, and indicate that the exploitation of both domain signatures of the clustered microcalcifications yields significantly better detection results.
Key words: wavelet transforms, high order statistics, tail ratio, synthetic microcalcifications.

Rachubi\'nski M.:
Iris Features Extraction Using Beamlets and Wedgelets.
MGV vol. 18, no. 3, 2009, pp. 289-304.
A new approach to iris feature extraction using geometrical wavelets is presented. Iris code is generated by using representation of the wavelet coefficients based on a wedgelet dictionary. The accuracy of identification in the case of head inclination by a certain angle for different ranges of possibilities of shifting the iris code is shown. Experimental results on the CASIA iris database show that the proposed method is effective and exhibits encouraging performance.
Key words: biometrics, iris recognition, personal identification, beamlets, wedgelets.

Yang B., Li S.:
Texture Classification Using Combined Image Decomposition Methods.
MGV vol. 18, no. 3, 2009, pp. 305 - 319.
The developments of multiresolution analysis, such as the wavelet, curvelet and contourlet transforms, have yielded adequate tools to characterize different scales of textures effectively. These methods exhibit different performances in processing texture images due to their different characteristics. In order to use those complementary characteristics simultaneously, a texture classification method by combining different image decomposition methods is proposed. The proposed method is compared with the methods where only one kind of multiresolution transform is used. The experimental results demonstrate that the combined features can effectively capture the complementary information from different image decomposition methods and obviously improve the texture classification accuracy.
Key words: Texture classification, multiresolution transformation, features extraction, combined features, support vector machine.

Hussein A. S.:
TA System for Reconstruction of Solid Models from Large Point Clouds.
MGV vol. 18, no. 3, 2009, pp. 321 - 344.
This paper presents an integrated system for reconstructing solid models capable of handling large-scale point clouds. The present system is based on new approaches to implicit surface fitting and polygonization. The surface fitting approach uses the Partition of Unity (POU) method associated with the Radial Basis Functions (RBFs) on a distributed computing environment to facilitate and speed up the surface fitting process from large-scale point clouds without any data reduction to preserve all of the surface details. Moreover, the implicit surface polygonization approach uses an innovative Adaptive Mesh Refinement (AMR) based method to adapt the polygonization process to geometric details of the surface. This method steers the volume sampling via a series of predefined optimization criteria. Then, the reconstructed surface is extracted from the adaptively sampled volume. The experimental results have demonstrated accurate reconstruction with scalable performance. In addition, the proposed system reaches more than 80% savings in the total reconstruction time for large datasets of O(107) points.
Key words:solid modeling, computational geometry, surface reconstruction, parallel implementation, Radial Basis Functions, CAD, point clouds.

Huber-Mörk R., Löhndorf M., Heiss-Czedik D., Mayer K., Penz H., Soukup D.:
Fast and Efficient Colour Inspection using Sets of Ellipsoidal Regions.
MGV vol. 18, no. 3, 2009, pp. 345-361.
This paper addresses inspection of d-dimensional data using ellipsoidal decision regions for problems in automated visual inspection. For the special case of d = 3, the proposed method is researched in detail with respect to efficient real-time operation and compact storage of ellipsoid parameters. In order to reduce storage requirements, a method based on clustering the parameters describing the shapes of ellipsoids is proposed. Results are presented for colour images used in the quality analysis step in banknote printing. Additionally, estimations of computational effort and storage requirements are provided.
Key words: industrial inspection, ellipsoidal regions, real-time image processing.

Ahmed K., El-Henawy I., Atwan A.:
Novel DWT Video Watermarking Schema.
MGV vol. 18, no. 3, 2009, pp. 363-380.
In this paper a new video watermarking scheme is proposed which depends on 2-level Discrete Wavelet Transform decomposition of each component of an RGB video frame. The scheme embeds independent watermarks into different shots. A genetic algorithm is employed to match shots to watermarks. The scheme chooses between the HL$_1$ of red or green or blue components of each frame based on a key and embeds error correcting code into one of them. The scheme is blind. Experimental results show that the scheme is robust against attacks such as frame dropping, frame averaging, frame swapping, statistical analysis, and MPEG-2 and MPEG-4 compression. The proposed scheme uses a composite three-element key to increase the security.
Key words: Video Watermarking, Discrete Wavelet Transform, Genetic, Shot, Frame.

Machine GRAPHICS & VISION, Vol. 18 (2009), No. 4:

Hiremath P.S., Prabhakar C.J.:
Symbolic Kernel Fisher Discriminant Method With a New RBF Kernel Function for Face Recognition.
MGV vol. 18, no. 4, 2009, pp. 383-403 .
In this paper, we present a new radial basis kernel function (RBF) in symbolic kernel Fisher discriminant analysis (symbolic KFD) to extract nonlinear interval type features for face recognition. The kernel-based methods form a powerful paradigm, they are not favorable to deal with the challenge of large datasets of faces. We propose to scale up training task based on the interval data concept. Our investigation aims at extending KFD to interval data using new RBF kernel function. We adapt symbolic KFD to extract interval type nonlinear discriminating features, which are robust enough to varying facial expression, viewpoint and illumination. In the classification phase, we employ the minimum distance classifier with the squared Euclidean distance measure. The new algorithm has been successfully tested using four databases, namely, the ORL face database, the Yale face database, the Yale face database B and the FERET face database. The experimental results show that the symbolic KFD with the new RBF kernel function yields improved performance.
Key words: Face Recognition, Symbolic Data Analysis, Kernel Fisher Discriminant Analysis, Interval Type Features, RBF kernel function.

Hazem M. El-Bakry:
New Fast Principal Component Analysis For Real-Time Face Detection.
MGV vol. 18, no. 4, 2009, pp. 405-425.
Principal component analysis (PCA) has various important applications, especially in pattern detection, such as face detection and recognition. In real-time applications, the response time must be as short as possible. In this paper, a new implementation of PCA for fast face detection is presented. Such implementation relies on performing cross-correlation in the frequency domain between the input image and eigenvectors (weights). Furthermore, this approach is developed to reduce the number of computation steps required by fast PCA. The "divide and conquer" principle is applied through image decomposition. Each image is divided into smaller-size sub-images, and then each of them is tested separately using a single fast PCA processor. In contrast to using only fast PCA, the speed-up ratio increases with the size of the input image when using fast PCA and image decomposition. Simulation results demonstrate that the proposed algorithm is faster than conventional PCA. Moreover, experimental results for different images show its good performance. The proposed fast PCA increases the speed of face detection, and at the same time does not affect the performance or detection rate.
Key words: Image Decomposition, Fast PCA, Cross correlation, Frequency Domain, Face Detection.

Degtyarev S.V., Miroshnichenko S.Yu., Titov V.S.:
Detection of Object Edges in Aerospatial Cartographic Images.
MGV vol. 18, no. 4, 2009, pp. 427-437.
Modern geographic information systems (GIS) have a wide range of applications, including town-planning, town traffic control, ecology, logistics and others. These GIS require frequent update's using information supplied by aerospatial cartographic images. The standard approach to processing aerospatial cartographic images is interactive vectorization using software tools. This paper presents a method for object edge detection which is part of automated vectorization of aerospatial cartographic images. The method uses edge detection with local scale estimation and is well suited for locating anthropogenic objects (such as buildings and roads). The effectiveness of the proposed method has been experimentally verified on real aerospatial cartographic images.
Key words: image processing, aerospatial images vectorization, edge detection, local scale estimation.

Okarma K., Lech P.:
Application of Monte Carlo Preliminary Image Analysis and Classification Method for Automatic Reservation of Parking Space.
MGV vol. 18, no. 4, 2009, pp. 439-452.
The paper presents a Monte Carlo-based method for preliminary image analysis and classification applied to parking space reservation. The proposed approach can be used especially for large parking areas with high density of parking places, where automatic recognition of some geometrical parameters of the entering vehicles can be helpful for optimised usage of the available parking area. After performing real-time classification of the entering vehicles, the shortest path towards the reserved parking place can be presented for the driver e.g. on some display screens. This requires appling a~fast classification method which can be easily implemented in embedded systems instead of some typical classification algorithms characterised by relatively high computational complexity. The Monte Carlo approach presented in the paper can be used as an efficient solution of this problem.
Key words: Monte Carlo method, statistical image analysis, automatic classification.

Martyn T.:
Exploring the infinite-time behavior of the chaos game: Approximation and interactive visualization of 3D IFSP and RIFS invariant measures using PC graphics accelerators.
MGV vol. 18, no. 4, 2009, pp. 453-476.
In this paper we tackle the problem of approximation and visualization of invariant measures arising from Iterated Function Systems with Probabilities (IFSP) and Recurrent Iterated Function Systems (RIFS) on $\Real^3$. The measures are generated during the evolution of a stochastic dynamical system, which is a random process commonly known as the chaos game. From the dynamical system viewpoint, an invariant measure gives a temporal information on the long-term behavior of the chaos game related to a given IFSP or RIFS. The non-negative number that the measure takes on for a given subset of space says how often the dynamical system visits that subset during the temporal evolution of the system as time tends to infinity. In order to approximate the measures, we propose a method of measure instancing that can be considered an analogue of object instancing for IFS attractors. Although the IFSP and RIFS invariant measures are generated by the long-term behavior of stochastic dynamical systems, measure instancing makes it possible to compute the value that the measure takes on for a given subset of space in a deterministic way at any accuracy required. To visualize the data obtained with the algorithm, we use direct volume rendering. To incorporate the global structure of invariant measures along with their local properties in an image, a modification of a shading model based on varying density emitters is used. We adapt the model to match the fractal measure context. Then we show how to implement the model on commodity graphics hardware using an approach that combines GPU-based direct volume raycasting and 3D texture slicing used in the object-aligned manner. By means of the presented techniques, visual exploration of 3D IFSP and RIFS measures can be carried out efficiently at interactive frame rates.
Key words: Fractals, Recurrent iterated function system, Invariant measure, Volume rendering, GPU programming.

Yuan T., Cheng W., Tao G., Lijun L.:
An Improved Interactive Color Image Segmentation Using Region-Based Graph Cuts.
MGV vol. 18, no. 4, 2009, pp. 477-488.
The problem of efficient interactive extraction of a foreground object in a complex environment is the primary one in image processing and computer vision. The segmentation method based on graph cuts has been studied over the recent years. There are two main drawbacks of these studies: decrease in performance when the foreground and the background have similar colors, and long computing time when the image is large. In this paper, we present a new foreground objects extraction method using a region-based graph cuts algorithm. The image is pre-segmented into a large number of small partitions using the mean shift (MS) method. We use the partitions to represent the nodes in the graph instead of pixels. This approach can reduce the optimization time, which is closely related to the number of nodes and edges in the graph. Compared with the pixel-based method, our method can yield an excellent performance and exhibit a faster speed.
Key words: Image Segmentation, Graph Cuts, Mean Shift (MS), Foreground, Interactive Extraction.

Reviewers' index
Authors' index
First Negative Selection Ratio, Final Acceptance Ratio
Contents of volume 18, 2009

17 (2008) main 19 (2010)