Research On The TDI CCD Signal To Noise Ratio And Charge Transfer Smear In Aerial Remote Sensing Applications

Charge coupled device is widely used in remote sensing, astronomical observationand night vision due to its merits of low noise, high quantum efficiency, widespectral response and high geometric stability. Time delay and integration mode ofCCD provides increased sensitivity without the sacrifice of spatial resolution, andthe effective integration time is increased by a factor of N, which is equal to thenumber of TDI stages. Therefore, it is used for light weight design of the aerialcameras in remote sensing applications.In the CCD-based cameras, there are several noises. In our analysis, the CCDnoises can be divided into five main categories: photon shot noise, dark current noise,photon response non uniformity (PRNU) noise, circuit readout noise andquantization noise. However, the time delay and integration function has an effect onthe noise representations when compared to the frame detectors. After studying thephoton electron conversion process thoroughly, we develop a series of noiseformulations with respect to the integration stages. Then we provide an universalTDI CCD camera signal to noise ratio (SNR) formulation considering the groundscene reflector ratio, system modulation transfer function (MTF), ect. Furthermore,we analyze the effect of the analog gain on the SNR.As we know, in the aerial remote sensing applications, the higher SNR does notalways means the high imaging quality due to the influence of aircraft forwardmotion and attitude variations. After researching on the operation of a typical TDICCD, we give a thorough understanding on causes of clocking smear. Then anelaborate mathematical model describing the charge transfer procedure is developed, and the modulation transfer function losses due to charge transfer is also presented,which shows that nearly one pixel smear will be introduced by traditional phasetiming. Therefore, we propose a novel charge transfer method, using which only1/2φ pixel smear will occur within the imaging operation of a single TDI stage,where φ represents the number of timing phases.