​

• Designed an intelligent scheduling system of taxi based on distribution, and fluctuation of supply and demand using video surveillance technology (with data from Shanghai Public Security Bureau)

• Proposed a taxi-identifying and classifying algorithm using Histogram of Oriented Gradient (HOG): training the sample database, extracting HOG feature, constructing vehicle classifier based on Support Vector Machine

• Optimized the algorithm specific to color detection and multi-angle recognition

• Constructed a platform for system implementation using OpenCV and MATLAB to extensively evaluate the system (received 96% positive class accuracy and 92% negative class accuracy in 1100 samples)