Bipul Neupane

Bipul Neupane
Bipul Neupane, 2023 Doctoral Academy Fellow

Earth observation and AI for urban building footprint extraction

A/Prof Jagannath Aryal and Prof Abbas Rajabifard

School / Faculty:
School of Electrical, Mechanical and Infrastructure Engineering
Faculty of Engineering and Information Technology

PhD Details

My PhD aims to push the frontiers of urban mapping research by crafting highly precise techniques for delineating building footprints (vector polygons) within fast-growing capital cities, leveraging Earth observation images and state-of-the-art intelligent methods.

Details to my PhD problem: The creation of precise urban base maps is a critical component of urban planning. Previously, these maps were meticulously hand-crafted. However, the rise of Artificial Intelligence (AI), particularly machine learning and deep learning, has revolutionised the method of identifying complex urban features such as buildings and roads from images. This modern approach automates the extraction of these features from Earth observation imagery (including satellite and airborne sensors), representing them as vector polygons and lines. Despite the advancements, evolving deep learning methods face challenges related to data quality, precision, and overall accuracy. These challenges hinder the robustness and reliability of urban feature datasets and methodologies, impacting data providers (both local and global) and policymakers. To overcome these obstacles, there is a pressing need to standardise data products and enhance the methods for increased reliability and robustness. Such improvements are essential for effectively supporting decision-making processes, which are pivotal in monitoring and managing sustainable urban infrastructure amidst rapidly growing populations. My PhD aims to develop highly accurate deep learning methods for extracting building footprints (vector polygons) in rapidly expanding capital cities from aerial imagery captured by Earth observation sensors, such as satellites.

Q & A

Why did you decide to do a PhD?
I enrolled into Geomatics Engineering for my undergraduate degree when I was a teenager. Much later, I began to love the idea behind Geographic Information System (GIS), the visually appealing cartographic maps, and the Earth observation data and tools to monitor and manage the Earth resources. Lucky, I enjoyed surveying and mapping the terrific geographies of the Himalayas, the place I was born. My interests grew into the emerging AI technologies that if combined with Earth observation, could change forever the methods of mapping the landscapes. After working for few years and getting a research-based Master’s degree with the majors on computer vision, my next step was clear for me – Research career on Earth observation and AI for accurate and automated mapping. Hence, the PhD.

What do you enjoy reading?
Mostly sci-fi and adventures. Occasionally, spiritual stories if they have a hint of adventure in it.

What do you enjoy doing when you're not working on your PhD?
I enjoy doing figurative arts and going for a long walk.

Name one fun fact about you.
I used to pet turtles and soon realised they don’t really enjoy being pet.

coming soon...