Selective Compression of Live Digital Video Frames for Power-friendly Wireless Transmission and Efficient Reconstruction

The project is aimed at implementing real-time processing techniques for video acquisition, compression and transmission. The project focuses on defining solutions in order to solve constraints within the system related to bandwidth and the battery energy of the sensor. Depending on the application, the acquisition and compression procedures could be based on extracting significant features and compressing them in a lossless fashion followed by data transmission over a wireless channel. The real-time transmission of data within the available bandwidth would be implemented so as to maximize the battery life while transmitting enough bits over the channel for an accurate image reconstruction. Specially tailored schemes for image recovery would be investigated such that the reconstructed image at the server and those streamed to the authorized end users would meet a certain high level quality. Furthermore, the algorithm would be automated to make it more adaptable under various conditions. These processes are relevant to the development of real-time machine vision sensor devices which have goals of energy conservation and ease of technological use.

Intern: 
Hiba Shahid
Faculty Supervisor: 
Dr. Rabab Ward
Province: 
British Columbia
Program: