A Compressed Sensing CMOS Image Sensor with Pixel-wise Coded Exposure (PCE)

Unmet Need:
This technology allows for independent control of exposure time during pixel integration whereby the image sensor can sample a high frame rate video at a low framerate.  
 
Technology Overview
Through modulating pixel exposure, a video is compressed into a single image for storage and transmission. The video is recovered from the image using compressed sensing techniques. This method allows for video compression prior to sampling. Further, in this architecture, the pixels have the ability to be exposed over multiple frames, thereby increasing video SNR as more light can be integrated onto the pixel. Researchers at Johns Hopkins have also increased signal-to-noise ratio and have avoided motion blurring effects due to long exposure time. Additional features of the technology include:
1. An all-CMOS implementation without using any optical apparatus for single pixel exposure control;
2. An integrated random exposure generation block that generates spatio-temporal exposure locally, enabling both high spatial and temporal resolution;
3. The ability to capture High Dynamic Range (HDR) scene using single image sensor
 
Stage of Development
Prototype

Publications
https://www.ncbi.nlm.nih.gov/pubmed/27137331
 

Category(s):