From the Publisher: 
The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and illustrated in pseudo code. Appropriate for use by engineers as a comprehensive reference to the computer vision enterprise.

Computer vision : a modern approach = 计算机视觉 : 一种现代的方法

MUCKE aims to mine a large volume of images, to structure them conceptually and to use this conceptual structuring in order to improve large-scale image retrieval. The last decade witnessed important progress concerning low-level image representations. However, there are a number problems which need to be solved in order to unleash the full potential of image mining in applications. The central problem with low-level representations is the mismatch between them and the human interpretation of image content. This problem can be instantiated, for instance, by the incapability of existing descriptors to capture spatial relationships between the concepts represented or by their incapability to convey an explanation of why two images are similar in a content-based image retrieval framework. We start by assessing existing local descriptors for image classification and by proposing to use co-occurrence matrices to better capture spatial relationships in images. The main focus in MUCKE is on cleaning large scale Web image corpora and on proposing image representations which are closer to the human interpretation of images. Consequently, we introduce methods which tackle these two problems and compare results to state of the art methods. Note: some aspects of this deliverable are withheld at this time as they are pending review. Please contact the authors for a preview.

http://ieeexplore.ieee.org/iel2/4524/12820/00592460.pdf

Image Processing

IEEE Transactions on Pattern Analysis and Machine Intelligence

Real and Complex Analysis

Image change detectio algorithms : A systematic survey

In most laser welding applications the welding result is highly sensitive to the weld position. The initial determination of process parameters therefore requires time-consuming adjustments to be made. When manufacturing individual parts with low quantities, this is a significant cost driver. State-of-the-art computer-aided design/computer-aided manufacturing systems ease the creation of weld seams significantly, but position deviations resulting, e.g., from workpiece tolerances or displacement must still be compensated. Seam tracking systems are rather complicated to handle during process qualification as they require adjustment to a start position and are subject to be misguided by disturbances in the material. Besides, it is difficult to apply seam tracking to scanner-based microwelding operations, which are considered in this paper. Therefore, the authors developed a system that enables the user to position welds easily and accurately by marking the intended weld position in a calibrated live image of the workpiece, which serves as reference. Thereby, positioning errors are compensated. When executing the welding process it is also possible to observe it in the live image and to assess the geometrical properties of the weld surface immediately. Moreover, the paper describes how this idea is extended to a system for creating more complex weld paths and concludes with an approach for automated image-based path generation by extracting the joint geometry from the workpiece image.

Image-based creation and positioning of welds for accelerated prototyping

Bei der Harnblasenvideoendoskopie ist das Navigieren und das kontrollierte Sichten des gesamten Hohlorgans durch das begrenzte Sichtfeld des Endoskops fur den Urologen stark erschwert. Als Navigationshilfe konnen hierbei zusatzliche externe Navigationssysteme eingesetzt werden. Da jedoch aufgrund zu hoher Kosten optische und magnetische System bisher nicht verwendet wurden, steigt das Interesse an low.cost Systemen. Hierzu wurde ein erstes zystoskopisches Navigationssystem mittels Tragheitssensorik entwickelt, welches mit einer Fehlertoleranz von ca. 2° Winkelgrad eine erste robuste und kostengunstige Orientierungshilfe fur die Harnblasenendoskopie darstellt.

/pdf/tragheitsbasiertes-navigationssystem-fur-die-2v58h8soib.pdf

Trägheitsbasiertes Navigationssystem für die Harnblasenendoskopie

PROBLEM TO BE SOLVED: To obtain optimum picture quality by multiplying the change of a picture value by a coefficient based on an average picture value in a window in the periphery of an objective picture element in an original picture and/or the dispersion of the picture value and the picture value itself. SOLUTION: The picture values of respective picture elements in a detailed picture Si are converted in a block 21 having a non-linear conversion characteristic. A conversion processing is executed by using a look-up table where the respective picture values of the detailed picture Si as input are related to the picture value of a conversion picture Sei which is possible as output. In a block 22, a difference is detected between the picture value as the output of the block 21 and the picture value of the objective picture element in the detailed picture Si. In a multiplication block 23, the difference is multiplied by the coefficient (b) of 0-1. The coefficient (b) is decided from the picture having a value matched with average lightness in an objective position on the respective picture elements of the detailed picture Si. At that time, the original picture Ii-1 of the detailed picture Si can be used since it has the totally similar number of picture elements as that of the detailed picture Si.

Picture processing method

The aim of the paper is to outline a layered statistical image model suitable for unsupervised image segmentation. The segment internal texture signal is described based on its spatial frequency representation while the image partition is modelled as a sample of a Gibbs/Markov random field. The most likely segmentation is estimated using a maximum a posteriori (MAP) formulation with the unknown parameters being determined by constrained least squares (CLS) estimation. >

A stochastic model for image segmentation involving constrained least squares estimation

Kurzfassung. In diesem Beitrag wird eine Methode zur echtzeitfahigen Block-Matching-basierten Bewegungsschatzung beschrieben, die auch in der Prasenz von starkem Rauschen zuverlassige Ergebnisse liefert. Zu diesem Zweck werden zunachst Blocke kreisformig im Bild verteilt. Mit Hilfe des Struktur-Tensors wird entschieden, ob ein Block genugend Strukturinformation aufweist, um zur Bewegungsschatzung herangezogen zu werden. Die ausgewahlten Blocke werden verwendet, um die Parameter eines globalen Bewegungsmodells unter Verwendung des RANSACAlgorithmus robust zu bestimmen.

/pdf/modellbasierte-echtzeit-bewegungsschatzung-in-der-25vcxo8mtb.pdf

Til Aach

Papers

Image-based creation and positioning of welds for accelerated prototyping

Trägheitsbasiertes Navigationssystem für die Harnblasenendoskopie

Picture processing method

A stochastic model for image segmentation involving constrained least squares estimation

Modellbasierte Echtzeit-Bewegungsschätzung in der Fluoreszenzendoskopie.