Big Data Technologies and Applications. 7th International Conference, BDTA 2016, Seoul, South Korea, November 17–18, 2016, Proceedings

Research Article

Archaeological Site Image Content Retrieval and Automated Generating Image Descriptions with Neural Network

Download
215 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-58967-1_9,
        author={Sathit Prasomphan},
        title={Archaeological Site Image Content Retrieval and Automated Generating Image Descriptions with Neural Network},
        proceedings={Big Data Technologies and Applications. 7th International Conference, BDTA  2016, Seoul, South Korea, November 17--18, 2016, Proceedings},
        proceedings_a={BDTA},
        year={2017},
        month={6},
        keywords={Image content retrieval Neural network SIFT algorithms Feature extraction},
        doi={10.1007/978-3-319-58967-1_9}
    }
    
  • Sathit Prasomphan
    Year: 2017
    Archaeological Site Image Content Retrieval and Automated Generating Image Descriptions with Neural Network
    BDTA
    Springer
    DOI: 10.1007/978-3-319-58967-1_9
Sathit Prasomphan1,*
  • 1: King Mongkut’s University of Technology North Bangkok
*Contact email: ssp.kmutnb@gmail.com

Abstract

This research presents a novel algorithms for generating descriptions of stupa image such as stupa era, stupa architecture by using key points generated from SIFT algorithms and learning stupa description from the generated key points with artificial neural network. Neural network was used for being the classifier for generating the description. We have presented a new approach to feature extraction based on analysis of key points and descriptors of an image. The experimental results for stupa image content generator was analyze by using the classification results of the proposed algorithms to classify era and architecture of the tested stupa image. To test the performance of the purposed algorithms, images from the well-known historical area in Thailand were used which are image dataset in Phra Nakhon Si Ayutta province, Sukhothai province and Bangkok. The confusion matrix of the proposed algorithms gives the accuracy 80.67%, 79.35% and 82.47% in Ayutthaya era, Sukhothai era and Rattanakosin era. Results show that the proposed technique can efficiently find the correct descriptions compared to using the traditional method.