This repository is used to maintain code used for Smart India Hackathon-2018, Software Edition.
Problem: Text Extractor, under ISRO, Ministry of Space (Code: ISR3) Problem Statement: Extract and Organize the regional text in the community assets Photographs -With an emphasis on transparency and accountability various ministries under Government of India have taken initiative of geo tagging the respective assets. While geo tagging the assets the latitude, longitude, photograph and other important attributes are being captured. -However, there is additional information of immense value embedded in the photographs in Regional Languages which are currently not used in decision making process due to lack of automated system for extraction & organizing the information embedded in the photograph. Desired Outcome -The Solution shall comprise an automatic (1) identification of the regional language. 3 Languages shall be attempted. English, Hindi & any one Regional Language of Team’s Choice (2) detection and extraction of text within the images (3) Organization of the extracted info. in a structured manner & Linking with the geographic coordinates Datasets https://nrsc.gov.in/hackathon2018/ Domain Computer Vision, Image to Text Conversion (E.G) Optical Character Recognition (OCR). Challenge Automatic language identification, extraction and Organizing the information Usage To enrich the Community Assets Inventory which is an important input in Developmental Planning User Public, Decision Makers and Panchayat Raj Institutions Expected No. of concurrent users Extraction will be scheduled offline on the allotted directory of photographs uploaded by the user, during geo-tagging. Role of user -To schedule the execution of the automated solution on the allotted directory of photographs uploaded by the user during geo-tagging Technicality Technical Available solutions (if yes, reason for not using them) Not Available Domain Expert(s) Vijaya Banu. A, Uday Raj.
Proposed Solution: https://drive.google.com/open?id=1P6iAuqnuCYXYmCrjRm-EN-Lkms1Yqv7x