Detta är ett uppsatsförslag hämtat från Nationella Exjobb-poolen. Klicka här för att komma tillbaka till samtliga exjobbsförslag.
Thesis Proposal: Pedestrian Detection for Volvo Technology Corporation
In this thesis, real-time image processing techniques will be developed to detect vulnerable road users (primarily pedestrians, but also considering cyclists etc).
The application of this work is to improve the safety of commercial vehicles with respect to vulnerable road users. It is often difficult for commercial vehicle drivers to clearly observe the entire region around the vehicle; hence some form of automatic assistance is desirable.
High level specifications of the system to be investigated include:
• Detect pedestrians from a camera mounted on a commercial vehicle, which is either stationary, or moving slowly
• Development in C++
• Hardware is a regular desktop PC
• System should operate offline on recorded videos, or even with live video feed (if time permits)
The task comprises the following:
• Reviewing state-of-the-art image classification and machine learning algorithms for pedestrian detection (Adaboost, SVM etc)
• Adapting/modifying existing algorithms to work on commercial vehicles
• Collecting image data to use for dector/classifier training
• Using C++ and open source software such as OpenCV, design and develop a prototype system running in real time
• Perform some investigations to evaluate performance of the system
Number of students:
Computer Science, Electrical Engineering or Engineering Physics; Knowledge of image processing and signal processing, C++, also real-time computing is very beneficial.
For further information, please contact
Grant Grubb, Systems Engineer, tel: +46 31 32 338 18
Frida Ramde, Group Manager, tel: +46 31 32 262 73
Please apply at www.volvogroup.com/career
Informationen om uppsatsförslag är hämtad från Nationella Exjobb-poolen.