projects | Yutian Lei

Main contributor

Building-User Linking integrating 5G Cellular Signaling Data and Spatio-temporal Behavior Analysis

We developed a two-stage approach that significantly improved building-level user positioning accuracy using 5G mobile signal data. The solution combines cellular signal metrics with users' spatio-temporal behavior patterns and building functionalities, effectively addressing the challenge of low-precision 5G MR data for accurate building-user linking.

Spatial Database System for Subgrade Compaction Monitoring with WebGIS Visualization (BSc. Thesis)

PostgreSQL PL/pgSQL WebGIS

I developed a real-time monitoring system that enables dynamic quality assessment of subgrade compaction through spatial analysis of compactor trajectories. The system features an innovative incremental update technique that improved geometric data processing efficiency from minutes to milliseconds, along with a web-based platform for real-time visualization and anomaly detection.

SmartTicket: A Geofencing-based Intelligent Ticketing System for Seamless Tourist Experience

JAVA PostgreSQL WebGIS

We developed an intelligent ticketing system that uses real-time BeiDou positioning and geofencing for automatic ticket verification, eliminating manual checks at scenic spots. The modular design ensures reliable performance and a seamless tourist experience. I contributed to API development for frontend-backend interaction and database design.

Participant

OGE - a Platform for Earth Observation, Spatial Data Processing and Analysis

JAVA PostgreSQL Database

I designed and implemented a PostgreSQL database to store global geospatial vector data, including database schemas, partitioning strategies, and table structures with appropriate fields and attributes.

An Agricultural Greenhouse Extraction Method Based on Sentinel-2 Data

Python PyTorch Remote Sensing

We proposed a novel approach combining Convolutional Neural Networks (CNN) with transfer learning, utilizing a ResNet34 network pre-trained on ImageNet to achieve high-precision identification of agricultural greenhouses using Sentinel-2 high-resolution imagery.