B.Tech- Electronics and Communication
National Institute of Technology, Kurukshetra
To excel in the field of research and innovation by employing my academic excellence, hard-work and dedication.
96% Marks, CBSE Board
10 CGPA, CBSE Board
Application Development for Royal Bank of Scotland
Worked on a web-service using .Net and Visual C# using SQL. The web-service was a vendor service for Infosys and was used on day-to-day basis for information retrieval. Also, developed a framework for automated reporting system using .NET MVC. The reporting system was customized with data to be sent weekly, monthly or yearly which was previously done manually. Gave a presentation on the deliverables towards the end of internship.
Sketch Based Image Recognition
Currently working on a user-interactive GUI for sketch-based image retrieval using the implementation of the 2016 IEEE Publication titled "Sketch Based Image Retrieval by Salient Contour Reinforcement". This involved algorithms like angular radial partitioning and boundary detection methods like Canny, Sobel. The paper discarded the outputs from such detection methods and used a "Berkeley Detector" as proposed in the paper "Learning to Detect Natural Image Boundaries Using Local Brightness, Color, and Texture Cues". The approach processed image in global and salient image regions. We are currently working to extend SBIR into applications like signature based document retrieval or animated video creator using sketches.
Virtual Paint using OpenCV
A project based on color detection using open source computer vision OpenCV library on the visual studio platform and C++. It focused on painting on screen from a specified distance, without any actual contact with the screen and also, without using a mouse pointer. This was done using color detection techniques by thresholding the image for specific color values. A colored bob had to be calibrated according to it's color and lighting conditions of the room using a control window. After calibration, the central point of the thresholded bob was obtained as the reference to its position on the screen. Major disadvantage of this project was that it was not immune to background noises or objects with similar color (HSV values) as that of the bob.