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

Fluorescence endoscopy is an emerging technique for the detection of bladder cancer. A marker substance is brought into the patient's bladder which accumulates at cancer tissue. If a suitable narrow band light source is used for illumination, a red fluorescence of the marker substance is observable. Because of the low fluorescence photon count and because of the narrow band light source, only a small amount of light is detected by the camera's CCD sensor. This, in turn, leads to strong noise in the recorded video sequence. To overcome this problem, we apply a temporal recursive filter to the video sequence. The derivation of a filter function is presented, which leads to an optimal filter in the minimum mean square error sense. The algorithm is implemented as plug-in for the real-time capable clinical demonstrator platform RealTimeFrame and it is capable to process color videos with a resolution of 768×576 pixels at 50 frames per second.

/pdf/denoising-fluorescence-endoscopy-a-motion-compensated-391v550ldo.pdf

Denoising fluorescence endoscopy - A motion compensated temporal recursive video filter with an optimal minimum mean square error parameterization

Zur Verringerung des Rauschen in mit niedriger Dosis aufgenommen Rontgenbildern werden von uberlappenden Bildblocken lokale Spektren bestimmt und der Betrag jedes Koeffizienten in Abhangigkeit vom momentanen Signal-zu-Rausch-Verhaltnis unterschiedlich stark abgeschwacht. Um hierbei Datenvermehrung zu vermeiden, werden kritisch abtastende Transformationen, d.h. die Lapped Orthogonal Transform (LOT) und die Modulated Lapped Transform (MLT) erfolgreich eingesetzt. Sowohl der visuelle Vergleich als auch die quantitative Auswertung ergeben, das insbesondere die MLT trotz deutlich niedrigerer Rechenzeit insgesamt den bisher verwendeten Transformationen DFT und DCT nicht unterlegen ist.

Spektrale Betragsschätzung als Methode zur Rauschunterdrückung in Röntgenbildern

In minimal invasive surgery (MIS) a complete and seamless inspection of organs, e.g. the urinary bladder, using video endoscopes is often required for diagnostics. Since the endoscope is usually guided by free-hand, it is difficult to ensure a sequence of seamless frame transitions. Also 2-D panoramic images showing an extended field of view (FOV) do not provide always reliable results, since their interpretations are limited by potentially strong geometric distortions. To overcome these limitations and provide a direct verification method, we develop a gap detection algorithm using graphs. Exploiting the motion information of the applied zig-zag scan, we construct a graph representation of the video sequence. Without any explicit global image visualization our graph search algorithm identifies reliably frame discontinuities, which would lead to holes and slit artifacts in a panoramic view. The algorithm shows high detection rates and provides a fast method to verify frame discontinuities in the whole video sequence. Missed regions are highlighted by local image compositions which can be displayed during the intervention for assistance and inspection control.

/pdf/gap-detection-in-endoscopic-video-sequences-using-graphs-tr2ei35bf3.pdf

Gap detection in endoscopic video sequences using graphs

Pleuramesothelioma is malignant tumor on the pleura, caused by asbestos exposure. Computer-assisted diagnosis system shall support physicians to assess the growth rate of detected pleural thickenings from CT data. This paper describes a new, improved method to automatically assess the size of detected thickening using thin plate spline interpolation which then leads to the 3D modeling of thickening. First, we detect each pleural thickening. Second, an automatic coordinate transformation is applied to enable the numeric calculation of the spline. Next, the appropriate landmark points are then automatically selected by using the newly applied chain-code concavity analysis and used as the constraint points for the interpolation. Numerical integration is applied to calculate area by slice. In the final step, the spline interpolation between layers is applied to calculate volume of each thickening. The results show that the thin plate spline interpolated boundary is suitable for 3D modeling of the thickening.

Automatic 3D modeling of pleural thickening through thin plate spline interpolation

Mit zytopathologischen Methoden kann Krebs sehr fruh schon anhand geringfugiger Normabweichungen in einzelnen Zellen erkannt werden. Nachteil ist oft der derzeit benotigte Zeitaufwand eines Zytopathologen, der insbesondere einen Einsatz als Screening-Verfahren ausschliesst. Eine wichtige zytologische Diagnosemethode ist die DNS-Bildzytometrie. Um deren Zeitaufwand zu reduzieren, mussen fur die Messung relevante Zellen automatisch detektiert werden. Fur Epithelien der Mundschleimhaut vergleichen wir daher verschiedene Varianten des k-Nachste-Nachbarn-Klassifikators (kNN), um zwischen sicher gesunden und krebsverdachtigen Zellbildern zu unterscheiden. Geeignete Merkmalskombinationen wurden durch das Floating-Search-Verfahren ausgewahlt.

/pdf/chromatinmuster-basierte-zellklassifizierung-fur-die-dns-4pdc6m8yyp.pdf

Til Aach

Papers

Denoising fluorescence endoscopy - A motion compensated temporal recursive video filter with an optimal minimum mean square error parameterization

Spektrale Betragsschätzung als Methode zur Rauschunterdrückung in Röntgenbildern

Gap detection in endoscopic video sequences using graphs

Automatic 3D modeling of pleural thickening through thin plate spline interpolation

Chromatinmuster-basierte Zellklassifizierung für die DNS-Bildzytometrie an Mundschleimhaut-Abstrichen