Sponsored Projects

  1. Visual Attention Assisted Image and Video Compression

    2. Project Code: VIV_ICG_2017_SGCIR
    No: IIT/SRIC/CS/VIV_ICG_2017_SGCIR/2018-19/085 Dated: 25-07-2018
    Sponsor: MHRD
    Duration: 15-07-2018 - 14-07-2021
    PI: Prof. Jayanta Mukhopadhyay, Co-PI: Prof. Partha Bhowmick

  2. Development of OCR and PDF viewer

    3. Project Code: OAP
    No. IIT/SRIC/CSE/OAP/2014-15/166 Dt. 14-10-2014.
    Sponsor: ARC Document Solutions India Pvt. Ltd., EM-4/1, 7th Floor, Sector V, Salt Lake, Techno India College Campus, Kolkata 700091
    Duration: 29-09-2014 to 28-09-2018 (48 months)
    PI: Professor Jayanta Mukhopadhyay, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Co-PI: Dr. Partha Bhowmick, Associate Professor, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Project summary:
    1. Developing an OCR engine that can perform accurate character recognition in real-time scenarios.
    2. Developing an image search engine for searching images within PDF documents (primarily engineering drawings).
    3. Providing a viewer for displaying AEC drawings in PDF format.

  3. Image analysis for preservation and archiving of Indian Cultural Heritage

    4. Project Code: APA
    No. NRDMS/11/1586/2009 Dt. 03-06-2010.
    Sponsor: DST, GOI
    Duration: August 2010 - July 2013
    PI: Professor Jayanta Mukhopadhyay, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Co-PI: Dr. Partha Bhowmick, Assistant Professor, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Project summary: Image analysis tools related to preservation and archival of Indian Heritage will be developed. In particular, attention will be paid in analysis of images of stone inscriptions and stone carving on the walls of monuments.

    In Hampi, there is an archeological museum where such stone carving and inscriptions are displayed. Also from the sites, the digital images of many such frescoes on the walls of monuments will be captured. The project will aim at suitable imaging of such samples and also developing processing tools, such as enhancement, segmentation, super-resolution operations, etc. for providing better perception of them. Effort will be made also to analyze and recover the missing parts in such images by exploiting contextual information. The system will also be facilitated with archiving and indexing of these images. Toward this an information system for annotation, storage and retrieval of these images will be developed. It will have a web-interface for disseminating information through internet.


  4. 3-D Image Sensor for Capturing Minute Surface Details and Visualization by Geometric Modeling

    5. Sponsor: DST
    Duration: 2010 - 2013
    PI: Dr. P. K. Biswas, Professor, Department of Electronics and Electrical Communication Engineering, Indian Institute of Technology, Kharagpur.
    Co-PI: Dr. Partha Bhowmick, Assistant Professor, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Project summary: A 3D image sensor for capturing fine details of depth profile will be developed. Attention will be paid in capturing the 3D surface profiles of stone inscriptions and stone carving on the walls of monuments. The art of carving stone has made its influence felt all over the world.

    In Hampi, there is an archaeological museum where such stone carving and inscriptions are displayed. Also from the sites, the digital 3D images of many such frescoes on the walls of monuments will be captured. The project will aim at suitable 3-D imaging of such samples in patches and also developing processing tools for registration and aggregation of such patches for providing better perception of them.

    Apart from 3D digitized data acquisition, the project will cover surface representation (triangulations, rendering, etc.). To have a successful and realistic output of the reconstructed output, visualization of the reconstructed model using state-of-the-art technologies would be the ultimate concern. The 3D image sensor once developed will be useful not only in this project but also in other application areas where capturing 3D depth profile is required.


  5. Preprocessing and Analysis of Degraded Documents

    6. Project Code: DRD
    Sponsor: Ministry of Communication & Technology
    SRIC Approval No. DRD/2011-2014/143 Dt. 25-01-2011
    Duration 3 years [wef 01-02-2011]
    PI: Professor Jayanta Mukhopadhyay, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Co-PI: 1. Dr. Shamik Sural, Associate Professor, School of Information Technology, Indian Institute of Technology, Kharagpur.
    2. Dr. Partha Bhowmick, Assistant Professor, Computer Science and Engineering Department, Indian Institute of Technology, Kharagpur.
    Project summary: Preprocessing of scanned documents is required to make them suitable for intermediate and higher level processing of different application, which may involve feature extraction, pattern recognition, semantic representation, etc. Noise removal is one such important preprocessing task. Due to different limitation of scanning procedure and poor quality of the physical source document, different types of noises are introduced in the digital images. For example, one such common phenomenon is the marginal noise, which may occur in the form of a large and dark region around the margins of document images, while scanning thick or skewed documents. There are other sources of noise such as ink-blobs, stray marks, etc. One of the objectives of this project is to develop algorithms for removal of different types of noise. Our approach will be mostly based on geometric processing and non-linear (and linear) filtering of images.

    Sometimes the scanned documents suffer from low-contrast and fading of colors. The proposed project aims at developing algorithms for contrast enhancement and color restoration of these documents. Our approach will be based on selective emphasis of factors responsible for expressing key features of a document. We would also use different color processing techniques such as saturation and de-saturation for color rendition, color constancy, etc.

    Color harmonization makes a document visually appealing by selection of proper combination of background and foreground colors of a document text and graphics. This is particularly demanding in presence of color images in a mixed document. The presence of different hues in the image looks for proper combination of colors in its text and graphics segments. The project intends to develop techniques for this type of processing. The later task involves also development of segmentation algorithms for a document with mixed content, such as text, graphics, and images.


  6. Strategic Aircraft Guidance

    7. Project Code: SAG
    Sponsor: ADA, Bangalore
    Duration 2007 - 2009
    PI: Professor Amit Kumar Das, Computer Science and Technology Department, BESU, Shibpur.
    Co-PI: Dr. Partha Bhowmick, Lecturer, Computer Science and Technology Department, BESU, Shibpur.
    Project summary: Feasibility study for collision avoidance algorithm for one-on-ten (Pursuer and Evader aircrafts) scenario. The scope includes: 1. Literature review on methods of collision avoidance between moving vehicles. 2. Methodology for generation of multi target trajectories in a combat situation with the purpose of collision avoidance assuming simultaneous visibility by any of the vehicles for the rest (position and velocity). 3. Refinement of the preliminary model to match the near exact situation with limits of Latax of 2G and the longitudinal acceleration of +/- 0.2G for the interceptors.