@article{2012,
  abstract     = {{The rapid growth of optical imaging technologies increased the access and collection of data, which boosts the demand of data and knowledge discovery. This is a fast growing topic in several industry and research areas. Nowadays, a large number of images and signals must be analysed in order to gain and learn proper knowledge. Detecting images with similar contents without specifying an image, recently attracts the researches in image processing domain. Motif discovery in image processing aims to tackle the problem of deriving structures or detecting regularities in image databases. Most of the motif discovery methods solve this problem by converting images into one dimensional time series in a pre-processing step and then applying a motif discovery on these one dimensional time series for image motifs detection. Nevertheless, this conversion might lead to information loss and also the problem of inability to discover shifted and multi-scale image motifs of different size. Contrary to other approaches, here, a method is proposed to find image motifs of different size in image data sets by employing images in original dimension (2D) without converting them to one dimensional time series.
The proposed approach consists of three steps: Mapping or transformation, feature extraction and measuring similarities. First, images are inspected by the Complex Quad Tree Wavelet Packet transform, which provides broad frequency analysis of an image in various scales. Next, statistical features are extracted from the wavelet coefficients. Finally, image motifs are detected by measuring the similarity of the features applying various similarity measures. Here, the performance of six similarity measures are benchmarked in details. Moreover, the efficiency of the proposed method is demonstrated on a data set with images from diverse applications such as hand gesture, text recognition, leaf and plant identification, etc. Additionally, the robustness of this method is examined with the image data overlaying with distortions such as noise and blur.}},
  author       = {{Deppe, Sahar and Lohweg, Volker}},
  issn         = {{1942-2679}},
  journal      = {{International Journal On Advances in Intelligent Systems}},
  keywords     = {{processing, Wavelet transformation.}},
  number       = {{3/4}},
  pages        = {{434 -- 446}},
  publisher    = {{IARIA Journals}},
  title        = {{{Evaluation of Similarity Measures for Shift-Invariant Image Motif Discovery}}},
  volume       = {{10}},
  year         = {{2017}},
}

@inproceedings{2022,
  abstract     = {{Nowadays, the boost of optical imaging technologies results in more data with a faster rate are being collected. Consequently, data and knowledge discovery science has become an attractive and a fast growing topic in several industry and research area. Motif discovery in image processing aims to tackle the problem of deriving structures or detecting regularities in image databases. Most of the motif discovery methods first convert images into time series and then attempt to find motifs in such data. This might lead to information loss and also the problem of inability to detect shifted and multi-scale image motifs  of different size. Here, a method is proposed to find image motifs of different size in image datasets by applying images in original dimension without converting them to time series. Images are inspected by the Complex Quad Tree Wavelet Packet transform which provides broad frequency analysis of an image in various scales. Next, features are extracted from the wavelet coefficients. Finally, image motifs are detected by measuring the similarity of the features. The performance of the proposed method is demonstrated on a dataset with images from diverse applications, such as hand gesture, text recognition, leaf and plant identification, etc. }},
  author       = {{Deppe, Sahar and Lohweg, Volker}},
  booktitle    = {{PESARO 2017 The Seventh International Conference on Performance, Safety and Robustness in Complex Systems and Applications}},
  editor       = {{Leister, Wolfgang}},
  issn         = {{2308-3700}},
  keywords     = {{Motif discovery, Image processing, Wavelet transformation}},
  location     = {{Venice, Italy }},
  pages        = {{27--32}},
  publisher    = {{The Seventh International Conference on Performance, Safety and Robustness in Complex Systems and Applications; Special track MAIS: Machine Learning Algorithms in Image and Signal Processing}},
  title        = {{{Shift-Invariant Motif Discovery in Image Processing 'Best Paper Award'}}},
  year         = {{2017}},
}