No icon

Pixel Binning for High Dynamic Range Color Image Sensor Using Square Sampling Lattice

Pixel Binning for High Dynamic Range Color Image Sensor Using Square Sampling Lattice

Abstract:

We propose a new pixel binning scheme for color image sensors. We minimized distortion caused by binning by requiring that the superpixels lie on a square sampling lattice. The proposed binning schemes achieve the equivalent of 4.42 times signal strength improvement with the image  resolution loss of 5 times, higher in noise performance and in resolution than the existing binning schemes. As a result, the proposed binning has considerably less artifacts and better noise performance compared with the existing binning schemes. In addition, we provide an extension to the proposed binning scheme for performing single-shot high dynamic range image acquisition.

Existing System:

Binning is a process of combining the electrical charges of multiple neighboring pixels into a single “superpixel”. The combined signal will be transformed into a digital signal by an analog-to-digital converter. This technique reduces the impact of read noise on the combined signal even if the individual pixel values are small. However, the noise performance improvements come at the price of spatial resolution loss. Binning was first introduced as a way to exploit CCD pixel sensor architecture. Today, a CMOS binning combines electrical charges by transistor switches that connect the photodiodes together. Binning has the advantage that it can be enabled only as needed. That is, the sensor may still operate at high resolution in ample light, and binning is used only when requiring better noise performance. Besides the improved readout circuits designs, the spatial sampling of combined pixels has significant impact on the binning performance. existing binning scheme such as Kodak PIXELUX and PhaseOne combine neighboring four pixels to yield an signal gain of 4 times. But our prior work in proved that the spatial and color artifacts caused by their binning patterns are due to aliasing stemming from the implied non-uniform sampling.

 

Proposed System:

We propose a new pixel binning design for color image sensors aimed at minimizing the resolution loss while improving noise performance. Our scheme combines neighboring pixels to yield superpixels arranged in a square sampling lattice. We achieve equivalence to 4.42 times signal strength improvement with image resolution loss of 5 times by ensuring that the binning pattern results in a square (i.e. uniform) sampling lattice. Contrast this to the existing binning scheme such as Kodak PIXELUX and PhaseOne that has a signal gain of 4 times but an effective resolution1 loss  of 16. As a result, we have considerably less spatial and color artifacts and better noise performance compared with the existing binning schemes. In addition, we provide an extension to the proposed binning scheme for performing a single-shot high dynamic range image acquisition.

Comment As:

Comment (0)