A Multi-Aperture Image Sensor With 0.7 $\mu{\hbox{m}}$ Pixels in 0.11 $\mu{\hbox{m}}$ CMOS Technology

Abstract

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> The first integrated multi-aperture image sensor is reported. It comprises a 166 <formula formulatype="inline"><tex Notation="TeX">$\times$</tex></formula> 76 array of 16 <formula formulatype="inline"><tex Notation="TeX">$\times$</tex> </formula> 16, 0.7 <formula formulatype="inline"><tex Notation="TeX">$\mu{\hbox{m}}$</tex> </formula> pixel, FT-CCD subarrays with local readout circuit, per-column 10-bit ADCs, and control circuits. The image sensor is fabricated in a 0.11 <formula formulatype="inline"><tex Notation="TeX">$\mu{\hbox{m}}$</tex></formula> CMOS process modified for buried channel charge transfer. Global snap shot image acquisition with CDS is performed at up to 15 fps with 0.15 V/lux-s responsivity, 3500 e- well capacity, 5 e- read noise, 33 e-/s dark signal, 57 dB dynamic range, and 35 dB peak SNR. When coupled with local optics, the multi-aperture image sensor captures overlapping views of the scene, which can be postprocessed to obtain both a high-resolution 2-D image and a depth map. Other benefits include the ability to image objects at close proximity to the sensor without the need for objective optics, achieve nearly complete color separation through a per-aperture color filter array, relax the requirements on the camera objective optics, and increase the tolerance to defective pixels. The multi-aperture architecture is also highly scalable, making it possible to increase pixel counts well beyond current levels. </para>