PhD Defense: 'A CMOS Frame Difference Event Sensor with Dynamic Range Extension'

The rapid growth of the Internet of Things is driving demand for low-power imaging systems tailored to machine-vision applications. Dynamic Vision Sensors are well suited for these environments due to their high temporal resolution, wide dynamic range, and low power consumption. Frame differencing provides another approach to reducing redundant data, lowering power usage, and simplifying storage and computation, and is widely used for motion detection and feature extraction. Although frame-difference cameras outperform conventional cameras in speed and power efficiency, they suffer from a significantly lower dynamic range and reduced responsiveness compared to Dynamic Vision Sensors. This PhD work aims to narrow that gap by enhancing the dynamic range of frame-difference sensors through the implementation of a lateral overflow capacitor. 

Supervisors:

• Víctor Manuel Brea Sánchez
• Andreas Kolb