Et profilbilde

Førsteamanuensis
Karl Skretting { "honorific-suffix": "Førsteamanuensis", "fn": "Karl Skretting", "tel": "Telefon: 51832016", "email": "karl.skretting@uis.no" }

Avdeling/enhet Det teknisk- naturvitenskapelige fakultet
Institutt/senter Institutt for data- og elektroteknikk
Rom KE E-422
Hjemmeside http://www.ux.uis.no/~karlsk/index.html
Tlf priv/mob 99489123

Forskningsområder

Dictionary learning and sparse approximation with applications in medical image processing and texture classification. Image processing in general, work has been done especially in image compression, see Image Compression Tools on the ICTools page. Now I am working on image regularization based on energy minimization principles.

Utvalgte publikasjoner

  • Recursive Least Squares Dictionary Learning Algorithm (Paper in pdf-format)
    • Karl Skretting and Kjersti Engan
    • IEEE Transactions on Signal Processing, pp 2121-2130, Vol. 58 No 4, April 2010
    • Digital Object Identifier: 10.1109/TSP.2010.2040671 
    • See also the Dictionary Learning Experiments, DLE, on its page.
  • A family of iterative LS-based dictionary learning algorithms, ILS-DLA, for sparse signal representation
    • Kjersti Engan, Karl Skretting and John Håkon Husøy
    • Digital Signal Processing, Volume 17, Issue 1, January 2007, Pages 32-49
    • Science Direct
  • Texture classification using sparse frame based representations (Draft in pdf-format)
    • Karl Skretting and John Håkon Husøy
    • EURASIP Journal on Applied Signal Processing, Volume 2006 (2006), Article ID 52561, 11 pages.
    • EURASIP JASP
  • General design algorithm for sparse frame expansions (Paper in pdf-format)
    • Karl Skretting, John Håkon Husøy and Sven Ole Aase
    • Elsevier Signal Processing, Volume 86, Issue 1, Pages 117-126, January 2006
    • Science Direct
  • Pågående forskning

    Dictionary learning for texture classification and image regularization based on energy minimization.

    Arbeidserfaring

    1986-1989 Block Watne AS
    1989-1996 Kverneland Underhaug AS
    1998-2002 Høgskolen i Stavanger, Ph.D. Student
    2002-present University of Stavanger, Associate professor

    Vitenskapelige publikasjoner (fra CRISTin)

    • Skretting, Karl (2002). Sparse signal representation using overlapping frames. NTNU. ISBN 8247154919. 161 s.
    • Barzideh, Faraz; Skretting, Karl; Engan, Kjersti (2015). The flexible signature dictionary. Eurasip; 2015-08-31 - 2015-09-04.
    • Colomer, Adrian; Naranjo, Valery; Engan, Kjersti; Skretting, Karl (2015). Retinal vessel inpainting using Recursive Least Square Dictionary Learning Algorithm. Eurasip; 2015-11-10 - 2015-11-13.
    • Skretting, Karl; Engan, Kjersti (2014). Energy Minimization by alpha-erosion for Supervised Texture Segmentation. AIMI (Association for Image and Machine Intelligence); 2014-10-22 - 2014-10-24.
    • Skretting, Karl; Engan, Kjersti (2011). Image Compression Using Learned Dictionaries by RLS-DLA and Compared with K-SVD. IEEE Singal Processing Society; 2011-05-22 - 2011-05-27.
    • Skretting, Karl; Engan, Kjersti (2011). Learned dictionaries for sparse image representation: Properties and results. SPIE - The International Society for Optical Engineering; 2011-08-21 - 2011-08-24.
    • Tengesdal, Morten; Rettedal, Arne; Skretting, Karl; Landsvik, Svein (2006). Studietilbod og FoU-aktivitetar ved Institutt for data- og elektroteknikk (Ide) ved Universitetet i Stavanger. 2006-09-28 - 2006-09-30.
    • Engan, Kjersti; Skretting, Karl; Husøy, John Håkon (2005). Denoising of Images Using Signal Dependent Frames and Matching Pursuit. 2005-03-18 - 2005-03-23.
    • Herredsvela, Jostein; Engan, Kjersti; Gulsrud, Thor Ole; Skretting, Karl (2005). Detection of lesions in mammograms using learned dictionaries and sparse representations. 2005-11-20 - 2005-11-25.
    • Herredsvela, Jostein; Engan, Kjersti; Gulsrud, Thor Ole; Skretting, Karl (2005). Detection of masses in mammograms by watershed segmentation and sparse representations using learned dictionaries. 2005-09-22 - 2005-09-24.
    • Herredsvela, Jostein; Engan, Kjersti; Gulsrud, Thor Ole; Skretting, Karl (2005). Texture Classification Using Sparse Representations by Learned Compound Dictionaries. 2005-11-16 - 2005-11-18.
    • Skretting, Karl; Engan, Kjersti; Husøy, John Håkon (2005). ECG compression using signal dependent frames and Matching Pursuit. 2005-03-18 - 2005-03-23.
    • Skretting, Karl; Husøy, John Håkon (2005). Frame based texture classification by considering various spatial neighborhoods. 2005-09-22 - 2005-09-24.
    • Husøy, John Håkon; Skretting, Karl (2003). A data oriented design theory for general signal expansions. 2003-09-10 - 2003-09-11.
    • Engan, Kjersti; Skretting, Karl (2002). A Novel Image Denoising Technique Using Overlapping Frames. 2002-09-09 - 2002-09-12.
    • Husøy, John Håkon; Skretting, Karl (2002). Texture classification using sparse frame based representations. 2002-10-04 - 2002-10-07.
    • Skretting, Karl; Husøy, John Håkon (2002). Partial search vector selection for sparse signal representation. 2002-10-02 - 2002-10-04.
    • Aase, Sven Ole; Skretting, Karl; Husøy, John Håkon; Engan, Kjersti (2001). Design of Signal Expansions for Sparse Representation.
    • Skretting, Karl; Engan, Kjersti; Husøy, John Håkon; Aase, Sven Ole (2001). Sparse representation of images using overlapping frames. 2001-06-11 - 2001-06-14.
    • Skretting, Karl; Husøy, John Håkon; Aase, Sven Ole (2001). A simple design of sparse signal representations using overlapping frames. 2001-06-19 - 2001-06-23.
    • Husøy, John Håkon; Aase, Sven Ole; Skretting, Karl; Engan, Kjersti (2000). Design of general block oriented expansions for efficient signal representation.
    • Skretting, Karl; Husøy, John Håkon; Aase, Sven Ole (1999). Improved Huffman coding using recursive splitting.
    • Herredsvela, Jostein; Engan, Kjersti; Gulsrud, Thor Ole; Skretting, Karl (2005). Detection of masses in mammograms by watershed segmentation and sparse representation using learned dictionaries.
    • Skretting, Karl; Husøy, John Håkon (2005). Frame based texture classification by considering various spatial neighborhoods.