A JOINT PIXEL AND REGION BASED MULTISCALE MARKOV RANDOM FIELD FOR IMAGE CLASSIFICATION

Article, Other literature type English OPEN
Mei, T. ; Zheng, L. ; Zhong, S. (2012)
  • Publisher: Copernicus Publications
  • Journal: (issn: 2194-9034, eissn: 2194-9034)
  • Related identifiers: doi: 10.5194/isprsarchives-XXXIX-B3-237-2012
  • Subject: TA1-2040 | T | TA1501-1820 | Applied optics. Photonics | Engineering (General). Civil engineering (General) | Technology
    arxiv: Computer Science::Computer Vision and Pattern Recognition
    acm: ComputingMethodologies_PATTERNRECOGNITION | ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION

MRF model is recognized as one of efficient tools for image classification. However, traditional MRF model prove to be limited for high resolution image classification. This paper presents a joint pixel and region based multi-scale MRF model for high resolution image classification. Based on initial image segmentation, the region shape information is integrated into MRF model to consider the pixel and region information simultaneously. The region shaped information is used to complement spectral signature for alleviating spectral signature ambiguity of different classes. The paper describes the unified multi-scale MRF model and classification algorithm. The qualitative and quantitative comparison with traditional MRF model demonstrates that the proposed method can improve the classification performance for regular shaped objects in high resolution image.
  • References (3)

    0.95 0.90 0.97 Zhang, L. and Q. Ji, 2010. Image Segmentation with a Unified Graphical Model. IEEE Transactions on Pattern Analysis and Machine Intelligence. 32(8), pP. 1406-1425.

    Zhen Lei, Tao Fang, and Deren Li, 2011. Land Cover Classification for Remote Sensing Imagery Using Conditional Texton Forest With Historical Land Cover Map. IEEE Geosci.

    Remote Sens. Lett., 8(4), pp.720-724.

  • Metrics
    No metrics available
Share - Bookmark