Analysis of Disparity Information for Depth Extraction Using CMOS Image Sensor with Offset Pixel Aperture Technique

Area of Science:

• Image Sensor Technology
• Computer Vision
• Optoelectronics

Background:

• Conventional cameras use external apertures on lenses, limiting integrated depth sensing.
• Complementary metal oxide semiconductor (CMOS) image sensors (CIS) are widely used in digital imaging.
• The offset pixel aperture (OPA) technique offers a novel approach to integrate aperture functionality within the sensor.

Purpose of the Study:

• To design and fabricate a CIS using the OPA technique for simultaneous depth and color information acquisition.
• To develop a geometric model for estimating disparity using the OPA technique.
• To validate the feasibility of depth extraction from a single CIS.

Main Methods:

• Fabrication of a CIS using a 0.11-µm CIS process incorporating left-offset (LOPA) and right-offset (ROPA) pixel apertures.
• Development of a color pattern comprising LOPA, ROPA, blue, red, and green pixels.
• Construction of a geometric model to estimate image disparity and comparison with measurement results.

Main Results:

• Simultaneous acquisition of disparity information from LOPA and ROPA channels.
• Simultaneous acquisition of both 2D color and disparity information.
• Depth extraction achieved using the single CIS with the OPA technique, with results comparable to estimations.

Conclusions:

• The OPA technique enables depth extraction from a single CIS, combining color and disparity information.
• This integrated approach simplifies depth sensing compared to conventional camera systems.
• The OPA technique is readily adaptable for commercial CIS cameras, offering potential for widespread adoption.