IMDSP-11.1

Modelling Magnetic Material Images with Simultaneous Autoregressions
Changjing Shang, D. M Titterington (Department of Statistics, University of Glasgow)

This paper presents a novel application of Simultaneous Autoregressive models to the synthesis of magnetic material images. The effect of using either symmetric or non-symmetric neighbour sets upon the visual and statistical properties of the resulting synthesised images are investigated. The use of a neighbour set whose shape corresponds to the orientations and coarseness of the texture allows the generation of synthetic images of good quality. Also, the size of such a neighbour set is usually smaller than that of the symmetric set required to reach similar modelling accuracy, thereby minimising the computational effort.

IMDSP-11.2

Multi-Hypothesis, Volumetric Reconstruction of 3-D Objects from Multiple Calibrated Camera Views
Peter Eisert, Eckehard Steinbach, Bernd Girod (Telecommunications Laboratory, University of Erlangen)

In this paper we present a volumetric method for the 3-D reconstruction of real world objects from multiple calibrated camera views. The representation of the objects is fully volume-based and no explicit surface description is needed. The approach is based on multi-hypothesis tests of the voxel model back-projected into the image planes. All camera views are incorporated in the reconstruction process simultaneously and no explicit data fusion is needed. In a first step each voxel of the viewing volume is filled with several color hypotheses originating from different camera views. This leads to an overcomplete representation of the 3-D object and each voxel typically contains multiple hypotheses. In a second step only those hypotheses remain in the voxels which are consistent with all camera views where the voxel is visible. Voxels without a valid hypothesis are considered to be transparent. The methodology of our approach combines the advantages of silhouette-based and image feature-based methods. Experimental results on real and synthetic image data show the excellent visual quality of the voxel-based 3-D reconstruction.

IMDSP-11.3

Infrared Sensor Modeling for Realistic Thermal Image Synthesis
Christelle Garnier (Ecole Louis de Broglie, D�partement informatique), Ren� Collorec (L.T.S.I., Universit� de Rennes I), Jihed Flifla (Ecole Louis de Broglie, D�partement informatique), Christophe Mouclier, Frank Rous�e (SOGITEC, Division �lectronique)

To generate realistic synthetic IR images, image acquisition by IR sensors must be reproduced. In this paper, we propose an IR sensor model based on physical laws governing effects involved in the IR imagery formation process. Our approach consists in a combination and an extension of current camera models used in visible and infrared image synthesis, and thus merges ray tracing and post-processing techniques. Our model represents the geometric and radiometric relationship between the points in the 3D observed scene and the corresponding pixels of the IR sensor output image. It offers the capability of simulating each IR sensor component in accordance with any given system technology and to any desired degree of precision. Moreover, it can also account for variations in many physical quantities through spatial, spectral, and temporal dimensions.

IMDSP-11.4

New Hardware-Efficient Algorithm and Architecture for the Computation of 2-D DCT on a Linear Systolic Array
Shen-Fu Hsiao, Wei-Ren Shiue (Inst. Comp. Eng., NSYSU, Taiwan)

A new recursive algorithm for fast computation of two-dimensional discrete cosine transforms (2-D DCT) is derived by converting the 2-D data matrices into 1-D vectors and then using different partition methods for the time and frequency indices. The algorithm first computes the 2-D complex DCT (2-D CCT) and then produces two 2-D DCT outputs simultaneously through a post-addition step. The decomposed form of the 2-D recursive algorithm looks very like a radix-4 FFT algorithm and is in particular suitable for VLSI implementation since the common entries in each row of the butterfly-like matrix are factored out in order to reduce the number of multipliers. A new linear systolic architecture is presented which leads to a hardware-efficient architectural design requiring only logN multipliers plus 3logN adders/subtractors for the computation of two NxN DCTs.

IMDSP-11.5

Estimating Crowd Density With Minkowski Fractal Dimension
Aparecido Nilceu Marana (UNESP - Rio Claro - SP - Brazil), Luciano da Fontoura Costa (USP - S�o Carlos - SP - Brazil), Roberto de Alencar Lotufo (UNICAMP - Campinas - SP - Brazil), Sergio A. Velastin (KCL - University of London - London - UK)

The estimation of the number of people in an area under surveillance is very important for the problem of crowd monitoring. When an area reaches an occupancy level greater than the designed one, people�s safety can be in danger. This paper describes a new technique for crowd density estimation based on Minkowski fractal dimension. Fractal dimension has been widely used to characterize data texture in a large number of physical and biological sciences. The results of our experiments show that fractal dimension can also be used to characterize levels of people congestion in images of crowds. The proposed technique is compared with a statistical and a spectral technique, in a test study of nearly 300 images of a specific area of the Liverpool Street Railway Station, London, UK. Results obtained in this test study are presented.

IMDSP-11.6

Data Efficient Implementation of UWBWA SAR Algorithms
Richard Rau, James H McClellan (Georgia Institute of Technology)

It is shown that the particular form of the frequency support of raw data and focused imagery obtained from an ultra-wideband, wide beamwidth synthetic aperture radar system can be exploited in non-separable sampling schemes to reduce the overall amount of raw data samples and image pixels that need to be stored and computed. Furthermore, it is demonstrated that the constant integration angle backprojection (CIAB) image former implicitly applies a fan filter that interpolates raw data sampled on a quincunx grid back onto the underlying rectangular grid. This subtle property of the CIAB has not been exploited so far. It leads to higher quality images with less computational complexity.

IMDSP-11.7

Supervised Classification for Synthetic Aperture Radar Image
Xavier Dupuis, Pierre Mathieu, Michel Barlaud (Laboratoire I3S Universit� de Nice-Sophia Antipolis)

This paper deals with the supervised classification of synthetic aperture radar (SAR) images. Our approach is based on two criteria, which explicitly take into account the intensity of the SAR image and the neighborhood classes, similary to the Pots model, but weighted by a discontinuity map. The high level of noise involves numerous classification errors, then we classify a restored image filtered with a well-adapted algorithm to clustering. Moreover, we isolate the texture of SAR images in order to help the classification. Finally, we present results on real SAR images.

IMDSP-11.8

Minimum Component Eigen-vector Based Classification Technique with Application to TM Images
Guohui He, Mita D Desai (University of Texas at San Antonio), Xiaoping Zhang

In this paper, we propose a new classification technique based on the Minimum Component Analysis (MCA) instead of the traditional Principal Components Analysis (PCA). Most existing classification techniques based on PCA represent a class by its principal component. However, the principal component is not always the best choice since there is a high possibility for classes to overlap with each other in the principal component direction. The new minimum component eigen-vector based classification technique overcomes this disadvantage by representing a class with its minimum component. In addition, a minimum likelihood decision rule is employed instead of maximum likelihood decision rule. Good performance of our technique is verified by experimental results on Kennedy Space Center (KSC) TM images.

< IMDSP-10 IMDSP-12 >

Last Update: February 4, 1999 Ingo Höntsch