Haobo Lv



NO.17 Xinxi Road

New Industrial Park

Xi'an Hi-Tech Industrial Development Zone


Xi'an Institute of Optics and Precision Mechanics of CAS







I am a 2nd year postgraduate student in the Center for OPTical IMagery Analysis and Learning (OPTIMAL), Chinese Academy of Sciences. My advisor is Prof. Yuan Yuan. I am interested in Remote Sensing, GIS, and Machine Learning.



~.         2012.9--now  


Signal and Information Processing

Chinese Academy of Sciences

~.         2008.9--2012.7  


Remote Sensing

Wuhan University(one of "985" and "211" program university)

Research Interests

~~         Remote Sensing: Dimensionality reduction, Change Detection, Atmospheric Remote Sensing

~~         Computer Vision: Visual Odometry, 3D scene parsing

~~         Machine Learning: Dimensionality Reduction, Semi-supervised Learning, Metric Learing

~~         2D & 3D scene parsing and its applications.

Research Experience

~         Research Project (2010.10-2012.4), Innovative Experiment of National Undergraduate Project of Wuhan University, Wuhan.

Mentor: [Liangming Liu], Juan Du, .

Project: Water Quality Monitoring of Liangzi Lake. Developed a system that uses HJ-1A/B staellite to monitor the quality of Liangzi Lake in Hubei.



Recent Projects

Data-dependent Semi-supervised Hyperspectral Image Classification

This paper proposed a Data-dependent semi-supervised (DDSS) method for hyperspectral image classification. This paper extends DDSS by considering unlabeled data to prove that the unlabeled data is helpful for DR, and the proposed can obtained a better performance..

A Semi-supervised Euclidean Embedding Dimensionality Reduction Method for Hyperspectral Image

This article proposes an approach to reduce the dimension of multivariate data, where a core contribution lies on the formality of including “unlabeled” samples in the proposed semi-supervised model for the task of data dimension reduction (DDR). Specifically, the approach employs Euclidean embedding for DDR and SVM to measure performance accuracy..



Semi-supervised Change Detection Method for Multi-temporal Hyperspectral Image

This project investigates novel metric learning methods to detect the landcover change information of multi-temporal hyperspectral imagery. We have developed efficient optimization methods for study the change areas, as well as learning methods for learning metric matrix. It was applied to three real hyperpsectral image applications and received good performance.

Screen shot 2013-07-22 at 8.47.15 PM

Water Quality Monitoring of Liangzi Lake by HJ-1A/B CCD Data

Innovative Experiment of National Undergraduate Project of Wuhan University, Wuhan. Developed a system that utilizes HJ-1A/B satellite to monitor the quality of Liangzi Lake in Hubei.



Haobo Lv, Y. Yuan, and L. xiaoqiang. Data-dependent Semi-supervised Hyperspectral Image Classification. In IEEE China Summit and International Conference on Signal and Information Processing . (ChinaSIP), 2013. [PDF]

D. Juan, Haobo Lv, Z. Yawen, L. Xiaojun,and Z. Xiang,. The water quality dynamic monitoring research based on HJ-1A/B CCD Data. The high resolution remote sensing data processing and application of workshop, ((ISTP)), 2012.

Friendly Links

[Lichao Mou] [Yong Yuan]