Real Time Mapping of a warehouse pillers using Omni-Directional Camera
Abstract: The main objective is to Map pillars in a Warehouse. In this project I was given set of real-time raw data, which includes sensor information like Omni Directional Images, time stamp and Position of vehicle from which the pictures were captured. With this sensor information I need to process the pictures and extract pillars and fuse it with vehicle position to predict the position of pillar in the real world. To achieve this Several algorithms were written in Matlab environment. Accomplished the results successfully. Results can be viewed slides below During this process I learned a lot of interesting real-time concepts from Prof. Björn Åstrand, who was my project supervisor. I should extended this project further to finish my Masters Thesis. But further objectives seams unrealistic to my situation at that time. I had to take a pause on this project due to other personal reasons.
This project work's few techniques can be compared to Google's Autonomous self driving car.