1. Summary

Research on similarity-based image retrieval based on the hierarchical model of image content elements.

2. Content-based Image Retrieval Based on the Hierarchical Model of Image Content Elements

Layer Content Elements Algorithms
5. Semantics Layer Parameters representing the semantics of the image or kansei (subjectivity, feeling) models of users evolutionary algorithms
4. Relation Layer Content elements representing relationship between regions Hierarchical attributed relational graph (image representation model)
3. Region Layer Content elements representing shape features or structural features of a region Shape decomposition (Representation of complex shape)
2. Pixel (Atom) Layer Class or label assigned through pixel-based classification / segmentation techniques Statistical pattern recognition
1. Observation Layer Image array obtained from the observation of the real world through sensors with the application of basic image correction techniques Satellite sensors, geometric correction

3. Demonstrations

  1. Digital Typhoon
  2. Digital Typhoon: Similarity-Based Image Search
  3. Digital Archive of Toyo Bunko Rare Books : SImilarity-Based Image Retrieval

4. References (Complete List)

  1. Asanobu KITAMOTO, "A Data Rearrangement Engine for Data Mining", Technical Report of IEICE (Institute of Electronics, Information, and Communication Engineers), Vol. PRMU2003-78, pp. 19-24, 2003-09 (in Japanese) [ Abstract ] [ PDF ]
  2. Asanobu KITAMOTO, Mikio TAKAGI, "Interactive Image Browsing Method Based on Evolutionary Computation", Proceedings of the 4th Symposium on Intelligent Information Media, pp. 173-180, 1998-12 (in Japanese) [ Abstract ] [ PDF ]
  3. Asanobu KITAMOTO, Mikio TAKAGI, "Interactive Image Browsing Using Queue-Based Genetic Algorithm", Journal of JSAI (Japanese Society for Artificial Intelligence), Vol. 13, No. 5, pp. 728-738, 1998-09 (in Japanese) [ Abstract ]
  4. Asanobu KITAMOTO, Mikio TAKAGI, "Hierarchical Model as a Framework for Constructing Similarity-Based Image Retrieval Systems", Technical Report of IEICE (Institute of Electronics, Information, and Communication Engineers), Vol. PRMU97-58, pp. 25-32, 1997-07 (in Japanese) [ Abstract ] [ PDF ]
  5. Asanobu KITAMOTO, "Similarity-Based Image Retrieval Using a Graph Structure Representing Regional and Spatial Information", Doctoral Dissertation, Electronic Engineering, Graduate School of Engineering, University of Tokyo, 1997-03 (in Japanese) [ Abstract ]
  6. Asanobu KITAMOTO, Mikio TAKAGI, "Retrieval of Satellite Cloud Imagery Based on Subjective Similarity", Proceedings of the 9th Scandinavian Conference on Image Analysis (SCIA'95), pp. 449-456, 1995-06 [ Abstract ] [ PDF ]