Mrinmoy Ghorai Homepage - Bishnupr Heritage Image Dataset

Bishnupur Heritage Image Dataset

Mrinmoy Ghorai* Pulak Purkait** Sanchayan Santra* Soumitra Samanta*** Bhabatosh Chanda*

*Indian Statistical Institute, Kolkata ** Toshiba Research Europe Ltd., Cambridge *** University of Liverpool, UK

Madanmohan_video.mp4

Abstract

Bishnupur is an attractive tourist place in West Bengal, India and is known for its terracotta temples. The place is one of the prospective candidates to be included in the list of UNESCO World Heritage sites. We intend to preserve this heritage site digitally and also to present some virtual interaction for the tourist and researchers. In this paper, we present an image dataset of different temples (namely, Jor-Bangla, Kalachand, Madan Mohan, Radha Madhav, Rasmancha, Shyamrai and Nandalal) in Bishnupur for evaluat-ing different types of computer vision and image processing algorithms (like 3D reconstruction, image inpainting, texture classification and content specific image retrieval). The dataset is captured using four different cameras with different parameter settings. Some datasets are extracted and earmarked for certain applications such as texture clas-sification, image inpainting and content specific image retrieval. Example results of baseline methods are also shown for these applications. Thus we evaluate the usefulness ofthis dataset. To the best of our knowledge, probably this is the first attempt of combined dataset for evaluating various types of problems for a heritage site in India.

Temples

More details about the temples and the dataset is available here.

3D Reconstruction

3D reconstruction of Jor Bangla and Nanda Lal datasets. (a),(f ) screenshot of sample images used for reconstruction displayed in VisualSFM toolbox; (b),(g) estimated camera poses and the sparse point-clouds; (c),(h) dense point-clouds estimated by CMVS/PMVS; (d),(i) corresponding line-meshes; (e),(j) 3D object of the texture maps. The reconstructions are near perfect and of high quality.

Image Inpainting

(a) Original image with damaged portion (b) masked image, and result of (c) Priority-based [4] (d) MRF-based [5] (e) Coherence-based [6].

Texture Classification

Texture image statistics in each group

Comparison

HTF - Haralick texture feature

SFTAF - Segmentation-based fractal texture analysis feature

HOG - Histograms of oriented gradients

BoVW - Bag-of-visual-words

FV - Fisher vector

DCNN - Deep convolution neural network

Content Specific Image Retrieval

Publication

"Bishnupur heritage image dataset (BHID): a resource for various computer vision applications" Mrinmoy Ghorai, Pulak Purkait, Sanchayan Santra, Soumitra Samanta and Bhabatosh Chanda Indian Conference on Computer Vision, Graphacis and Image Processing (ICVGIP), 2016, 80:1-80:8