Please use this identifier to cite or link to this item: https://scholarbank.nus.edu.sg/handle/10635/168817
Title: 3D OBJECT DETECTION AND POSE ESTIMATION USING POINT CLOUDS
Authors: SYEDA MARIAM AHMED
Keywords: 3D object detection, Edge-Aware PointNet, Bayesian Networks
Issue Date: 22-Aug-2019
Citation: SYEDA MARIAM AHMED (2019-08-22). 3D OBJECT DETECTION AND POSE ESTIMATION USING POINT CLOUDS. ScholarBank@NUS Repository.
Abstract: The digital revolution in image processing has integrated cameras into our daily lives, where image based algorithms are being used in all sorts of consumer products like mobile phones, cars, vacuum cleaners, traffic surveillance systems, etc. As the demand for higher accuracy and improved perception grows, there is a strong need for advanced algorithms that are useful for 3D data analysis and interpretation. This thesis studies the problem of 3D object detection for point clouds in indoor environments, focusing on heavily cluttered and occluded objects. While 3D deep learning networks are split between methods that leverage the mature 2D architectures to 3D or directly operate on unorganized points, there are significant challenges associated with both approaches. CNN based methods are inefficient due to sparsity in data while point based networks have scalability issues as the computational complexity of the network increases with the number of points processed. Consequently, the state-of-the-art performance for 3D object detection in indoor environments is much lower and requires a detailed analysis of the existing challenges.
URI: https://scholarbank.nus.edu.sg/handle/10635/168817
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
Syeda_Mariam_Ahmed_PHD_Thesis.pdf6 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

49
checked on Mar 5, 2021

Download(s)

15
checked on Mar 5, 2021

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.