Panqu Wang -- Research and Projects     Back to Main

 

 

Research Interest:

 

As an undergraduate, I have a broad interests focus on Digital Image Processing and Computer Vision. Now I put more emphasize on object recognition task using statistical learning method and discriminant saliency tools.

 

 

 

Research Projects:

 

Suspicious Object Recognition Method in Video Stream Based on Visual Attention

Advisor: Prof. Liming Zhang, Fudan University                          Status: Completed

 

We proposed a state-of-art method for intelligent object recognition and video surveillance based on human visual attention. Bottom-up and top-down attention are applied respectively in the process of acquiring interested object (saliency map) and object recognition. The revision of 4-channel PFT method is proposed in bottom-up attention, which enhances the speed and accuracy. Inhibit of return (IOR) is applied in judging the sequence of saliency object “pop-out”. Considering the basic characters of human vision, we also apply the rule that object looks small in the distance and big on the contrary. Euclidean distance of color distribution, object center coordinates and speed are considered in judging whether the target is match and “suspicious.” The extensive test on videos and images show that our method in video analysis has high accuracy and fast speed compared with traditional method. The method can be applied into many fields such as video surveillance and security.

 

 

See DEMO here

 

 

For details of the project, please refer to the Thesis(English Version | Chinese Simplified Version).

 

 

 

 

 

Real-time Automatic Object Detector Using Saliency

Advisor: Prof. Nuno Vasconcelos, University of California, San Diego          Status: Ongoing

 

Can we use the camera or cellphone to detect specific object automatically without being an expert in computer vision or pattern recognition? That’s what we do! To make the training set downloaded from Internet more specific and accurate, we use discriminant saliency algorithm to automatically crop the ROI area. The train is iterated by hierarchical EM method. After we obtain the positive and negative dataset, we start training the object detector using Adaptive ECBoost algorithm, which is more advanced than classical Viola&Jones algorithm. Now we are testing the detector on speed limit sign, cars, faces and many other categories. Hopefully, if the performance is high in speed and accuracy, we will embed the detector into a cellphone to auto-detect the specific object we need.

 

See Project Page HERE.

Demo:

 

 

 

 

 

 

 

 

Product Safely Based on Green Chemistry Integration Reaction Process- Real-time Detection and Control

Advisor: Prof. Qiyong Lu, Fudan University                             Status: Completed

 

Recent years, the wide application of electronic platform (embedded systems) has been successfully used in lots of other fields. In our project, the multidisplinary merge between Green Chemistry and Embedded System provides us a new approach to enhance the efficiency of chemical production and make the process much safer. We use LPC3230 to monitor the concentration change of Toluene diisocyanate(TDI) in real-time, draw the diagram, calculate the fitting accuracy and set up the kinetic model. We also use the interface of LPC3230 to connect iC-iR and Liquid Chromatograph together, make the whole process automatic.

 

 

 

 

 

Photovoltaic Power Generation Simulator

 

Advisor: Dr. Jingfei Jiang, Fudan University                             Status: Completed

The Photovoltaic Power Generation has become a hot topic in the world that seek for clean and renewable energy resources. As a project for the National Electrical Design Contest, we created the simulator in 4 days and 3 nights. Cored for SCM C8051F340 to generate SPWM wave, used PLL to sample and control, building DC-AC converter to produce higher-voltage sine wave. This system has the advantage of fast speed, high control efficiency, low load regulation and cost-effective.

 

 

 

For details, please refer to the THESIS.

 

 

 

 

 

Other Projects:

Besides the big projects related to my interested area, I also have a wide interest to all relevant fields of Electrical and Computer Engineering and did some very interesting projects with the faculties. Actually, I should thanks to every project I have completed- not only for the experimental skills gained, but they help me find what I really interest in under the immense knowledge ocean of ECE.

 

[Programming and System] Dynamic Storage Allocation                             Advisor: Prof. Bo Hu                          [PDF]                     Status: Finished

[Programming and Network] The Simulation of BA model               Advisor: Prof. Xiang Li                     [RAR]                     Status: Finished

[Biomedical Engineering] PACER Circuit Design                                          Advisor: Dr. Guohui Zhou              [PDF]                     Status: Finished

[Circuits and Systems] 6-Digit ADC System Design                                        Advisor: Prof. Yong Wang             [PDF]                     Status: Finished