ABSTRACT

With the goal of optimizing high resolution, low-noise, back-illuminated CCD night vision imaging performance within the context of commercial manufacturability and cost effectiveness, PixelVision, Inc., in partnership with Scientific Imaging Technologies, Inc. (SITe) has undertaken, using its commercial Sandbox^TM CCD processing service, a series of experimental design, performance optimization, and manufacturability efforts that has resulted in greater than 160 independent CCD designs. These low-light- level imagers have been demonstrated superior to conventional image intensified approaches for a broad range of night vision applications.

This paper describes a systems level approach by the authors to globally enhance all of the performance parameters of back-illuminated CCD technologies with the goal of manufacturing high yielding, cost effective imagers optimized for night vision applications. We group the CCD performance parameters (which for low light level imaging must be simultaneously optimized) into four primary CCD operations:

1. Charge Generation (e.g. QE),

2. Charge Collection (e.g. MTF),

3. Charge Transfer (e.g. CTE), and

4. Charge Detection (e.g. Read Noise).

To demonstrate the process used to optimize these four CCD operations for night vision applications, this paper contains CCD Transfer Curve data for two PixelVision Sandbox^TM CCD devices. CCD Transfer Curves, a formal methodology of CCD test and optimization, is the only demonstrated method of characterizing all of the performance characteristics of a CCD in a single, context independent data set.