Logo des Repositoriums

Auflistung nach Autor:in "Opitz, Felix"

1 - 10 von 11
  • Konferenzbeitrag
    The differential geometric view of statistics and estimation
    (Informatik 2009 – Im Focus das Leben, 2009) Opitz, Felix
  • Konferenzbeitrag
    Expectation maximisation for sensor data fusion
    (INFORMATIK 2006 – Informatik für Menschen, Band 1, 2006) Opitz, Felix
    The expectation maximisation algorithm offers several applications in sensor data fusion. An overview of some of this applications and a short course in expectation maximisation algorithm and its properties is given. The expectation maximisation algorithm (EM) was introduced by Dempster, Laird and Rubin in 1977 [DLR77]. The basic of expextation maximisation is maximum likelihood estimation (MLE). In modern sensor data fusion expectation maximisation becomes a substantial part in several applications, e.g. multi target tracking with probabilistic multi hypothesis tracking (PMHT), target extraction within probability hypothesis density (PHD) filter, cluster analysis within multidimensional data association, or image computing.
  • Konferenzbeitrag
    Fusion of IR/CCD video streams and digital terrain models for multi target tracking
    (Informatik 2009 – Im Focus das Leben, 2009) Dästner, Kaeye; Köhler, Bastian; Opitz, Felix
  • Konferenzbeitrag
    Information fusion: protection of military forces and homeland security
    (INFORMATIK 2010. Service Science – Neue Perspektiven für die Informatik. Band 2, 2010) Dästner, Kaeye; Köhler, Bastian; Opitz, Felix
    Information fusion is a key factor to enable Homeland security and military protection systems. Applications have to detect, track, classify and identify objects in real time. Sensor resources must be allocated to these objects of interest and the whole situation has to be analyzed concerning threats. Whenever necessary counter measures have to be initialized.
  • Konferenzbeitrag
    Markov random fields and the optical flow
    (INFORMATIK 2011 – Informatik schafft Communities, 2011) Opitz, Felix
    The optical flow can be viewed as the assignment problem between the pixels of consecutive video frames. The problem to determine the optical flow is addressed for many decades because of its central relevance. This paper gives a short resume about classical methods. Afterwards advanced Markov random fields are developed. The challenge and beauty of this approach consists of the large spectrum of mathematical and physical disciplines which work together.
  • Konferenzbeitrag
    Modern methods of data association
    (Informatik 2005 – Informatik Live! Band 2, 2005) Opitz, Felix; Kausch, Thomas
    Data association considers the relationship between the single sensor measurements and the original, real objects. Here, the special class of multidimensional data association is considered for multi-scan as well as unresolved measurement problems in multi-object tracking.
  • Konferenzbeitrag
    An Object Oriented Approach for Data Fusion
    (Informatik 2007 – Informatik trifft Logistik – Band 2, 2007) Dästner, Kaeye; Kausch, Thomas; Opitz, Felix
    An new object oriented development suite for data fusion is presented. It is shown how the various issues in the data fusion development like design, implementation, simulation and testing may be supported. This allows the realisation of high sophisticated data fusion systems as applied in numerous civil and defence areas: e.g. air traffic control, coastal surveillance, vehicle guidance systems, or safety and defence applications. The approach supports the design of multi layer data fusion systems with different architecture integrating numerous different information sources. Besides of the generation of an object oriented data fusion kernel the developer is also supported in rapid prototyping just at the beginning of the developing process. Further the final software and system testing and verification within a simulated and real environment is taken into account.
  • Konferenzbeitrag
    Reentry of space objects: Tracking and classification with sequential Monte Carlo techniques
    (Informatik 2005 – Informatik Live! Band 2, 2005) Opitz, Felix
    A new approach of tracking and classification of space objects with sequential Monte Carlo methods and numerical integration techniques is given.
  • Konferenzbeitrag
    Sensor integration in the security domain
    (Informatik 2009 – Im Focus das Leben, 2009) Köhler, Bastian; Opitz, Felix; Dästner, Kaeye; Kouemou, Guy
  • Konferenzbeitrag
    Sensor resolution models and multidimensional data association
    (INFORMATIK 2006 – Informatik für Menschen, Band 1, 2006) Opitz, Felix
    The multidimensional data association methods were developed to establish the relation between plots and tracks especially in dense target situations. However, even these advanced multidimensional data association methods lack in situations of unresolved measurement. Specifically, in real dense target situations being of most interest the phenomena of unresolvedplots happens quite often due to the limited sensor resolution.