Bingnan Li

Bingnan Li

Auto-generated profile, based on publicly available information

Bingnan Li

Civil and Environmental Engineering, University of New South Wales, Sydney, NSW 2052, Australia

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 February 2024 in International Journal of Applied Earth Observation and Geoinformation
Verified
Journal Article
Open access
BLBingnan Li
Bingnan Li
JGJiuchong Gao
Jiuchong Gao
SCShuiping Chen
Shuiping Chen
SLSamsung Lim
Samsung Lim
HJHai Jiang
Hai Jiang

11 May 2022 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
BLBingnan Li
Bingnan Li
JGJiuchong Gao
Jiuchong Gao
SCShuiping Chen
Shuiping Chen
SLSamsung Lim
Samsung Lim
HJHai Jiang
Hai Jiang
A point-of-interest (POI) represents a specific point location that may be useful or interesting for people, and therefore, each and every building footprint in a topographic map can be recognized as a POI. Automatic extraction of building footprints using remote sensing images has become a challenging and important research topic, which is in demand for urban planning and development. Extensive studies have explored a variety of semantic segmentation methods using deep learning algorithms to achieve better performance in building footprint extraction; however, the existing algorithms were shown to have some limitations, which lead to poor segmentation results. Building roofs were recognized as building footprints in the previous studies. This is prone to error especially for high-rise buildings due to different sensor view angles. In this article, we propose a multitask Res-U-Net model with an attention mechanism for the extraction of the building roofs and the whole building shapes from remote sensing images and then use an offset vector method to detect the footprints of the high-rise buildings based on the boundaries of the corresponding building roofs and shapes. We also apply the online food delivery (OFD) data to parse the POI name of every building footprint. Several strategies are also developed in combination with the proposed model, including data augmentation and postprocessing. We conduct numerical experiments using real data of remote sensing images and OFD historical order data. Results demonstrate that our proposed model achieves a total F1-score of 77.05% and an intersection over union (IoU) of 63.55% in terms of the building roof segmentation, and an overall F1-score of 79.02% and an IoU of 66.05% for the whole building shape segmentation, which both achieve the best performance among all baseline models.

2 October 2024 in Geo-spatial Information Science
Verified
Journal Article
Open access
BLBingnan Li
Bingnan Li
DHDanielle Hutchinson
Danielle Hutchinson
SLSamsung Lim
Samsung Lim
RMRaina MacIntyre
Raina MacIntyre


Verified
Journal Article
Open access
PDF file
GLGuannan Li
Guannan Li
GLGaohuan Lv
Gaohuan Lv
BLBingnan Li
Bingnan Li
XWXiang Wang
Xiang Wang
FZFen Zhao
Fen Zhao
Oil spills, whether in open water or near shorelines, cause serious environmental problems. Moreover, polarimetric synthetic-aperture radar provides abundant oil spill information with all-weather, day–night detection capability, but its use is limited by data usage and processing costs. Compact Polarimetric (CP) systems as a subsequent emerging system, which balance data volume and system design requirements, are promising in this regard. Herein, we utilize multisource oil spill scenarios and datasets from multiple polarimetric modes (VV-HH, π/4, DCP, and CTLR) to assess the oil spill detection capability of each mode under varying incidence angles conditions, spill causes, and oil types. Using qualitative and quantitative evaluation indicators, we compare the typical features of the multiple polarization modes as well as assess their consistency with Full Polarization (FP) information and their oil spill recognition performance across different incidence angles. In large-incidence-angle oil spill scenarios, the VV–HH mode exhibits the highest information consistency with the FP mode and the strongest oil spill recognition ability. At small incidence angles, the CP mode (i.e., CTLR mode) exhibits the best overall performance, benefiting from its effective self-calibration capability and low noise sensitivity. Furthermore, despite containing comprehensive information, the FP mode is not always superior to the dual-polarization and CP modes. Thus, in oil spill scenarios across different incidence angles, incorporating features from an appropriate polarization mode into oil spill information extraction and recognition can optimize the associated efficiency.