Title: Evaluating Features and Classifiers for Road Weather Condition Analysis

Reporter: Yiming Qian, PhD Candidate in Computer Science at York University, Canada

Time:  7:00PM-8:00PM, 24 April 2016

Address:5-207, Building No. 5, School of Remote Sensing and Information Engineering, Wuhan University



Weather-dependent road conditions are a major factor in many automobile incidents; computer vision algorithms for automatic classification of road conditions can thus be of great benefit. This paper presents a system for classification of road conditions from still-frames taken from an uncalibrated dashboard camera. The problem is challenging due to variability in camera placement, road layout, weather and illumination conditions. The system used a prior distribution of road pixel locations learned from training data then fuses normalized luminance and texture features probabilistically to categorize the segmented road surface. We attain an accuracy of 80% for binary classification (bare vs. snow/ice-covered) and 68% for 3 classes (dry vs. wet vs. snow/ice-covered) on a challenging dataset, suggesting that a useful system may be viable.


Yiming Qian is currently a Ph.D. student in Computer Science at York University, Canada specialized in texture classification, deep learning and statistical machine learning. He received his both Bachelor degree (Power Engineering) and Master degree (Electrical Engineering) from Ryerson University, Canada. Prior to his Ph.D study, from 2012 to 2015, he was a licensed R&D Engineer (P.ENG) in Siemens Energy Management high voltage instrumental transformer lab (Canada) specialized in electromagnetic field modelling, power system analysis and manufacturing process optimization. From 2011 – 2014, he was a part-time digital colour specialist in an industry printing technology startup – SMARTia Technologies Inc specialized in colour theory and printing.

Presented by Dr. Zhenguo Li