Digital pixel sensor array with logarithmic delta-sigma architecture

Area of Science:

• Electrical Engineering
• Computer Vision
• Sensor Technology

Background:

• Logarithmic image sensors offer wide dynamic range at video rates, mimicking the human eye.
• A key limitation of logarithmic sensors is their low peak signal-to-noise-and-distortion ratio (PSNDR).

Purpose of the Study:

• To enhance the PSNDR of logarithmic image sensors.
• To develop a novel pixel architecture integrating delta-sigma analog-to-digital converters (ADCs).

Main Methods:

• Designed and fabricated an image sensor with per-pixel delta-sigma ADCs using 0.18 micron CMOS technology.
• Implemented a small-area, low-power delta-sigma design for scalability.
• Included a decimator in each pixel for efficient data processing.

Main Results:

• The developed image sensor achieved a PSNDR superior to existing state-of-the-art logarithmic sensors.
• The sensor's PSNDR was found to be comparable to the performance of the human eye.
• Performance was benchmarked against various linear and nonlinear image sensors from industry and academia.

Conclusions:

• Integrating delta-sigma ADCs into each pixel is an effective method to improve the PSNDR of logarithmic image sensors.
• This approach offers a scalable solution for high dynamic range imaging with performance rivaling biological systems.
• The novel architecture represents a significant advancement in image sensor technology.