Novel Method To Upgrade The Low-light-level Signal Detection Sensibility

Low-light-level optical signal detection which includes the signal aquistion,conversion, enhancement, recording and display can expand human vision function inspace, time, and frequency domain. Low-light-level optical signal detection issignificant in science research, safety engineering and space astronomy. Sensitvity isan important parameter of measuring instrument.The instrument can sense tiny signalchange with higher sensitivity. To improve the instrument sensitivity is a baiscapproach in weak signal detection. This thesis provides a novel method to improve thelow-light-level optical signal detection sensivity. The principle of this method comesfrom the over-sampling technology of A/D conversion. By superposing a saw-toothwave light which we called optical shaped function on the low-light-level opticalsignal, we can get the unbiased estimate for the low-light-level optical signal aftersuperposition signal accumulation, thus improving the low-light-level optical signaldetection senstibility. The main contents of this thesis are as follows:(1) By analysis of its working principle, noise source, sensitivity for thephotoelectric direct detection system, we construct the black-box model for thephotoelectric direct detection system. Based on this model, the photoelectric directdetection system can be considered as an AD conversion system. We incorporate theoversampling and shaped function technique of AD conversion into the detection,therefore the detection sensitivity of the photoelectric direct detection system has beenimproved. After a preliminary principle analysis, we prove its efficacy by theexperiment with a direct detection system for the1-D low-light-level optical signaldetection.(2) Based on the analysis of1-D low-light-level optical signal detection, a novelmethod which employs the frame accumulation and shaped function is given toimprove the low-light-level image signal detection sensibility. By using theprobabilistic theory, we demonstrate that we can get the unbiased estimate for thelow-light-level image with this method. According to the Lindberg-Levy Theory, wecalculate out the image sensor’s SNR (signal-to-noise ratio) which has been improvedm△x_s~2/[sx_s (1-△x_s)] times than before. The efficacy of this method has been proved by alow-light-level shadow image detection experiment. (3) The method which employs the frame accumulation and shaped functiontechnique is effective in low-light-level imaging. Yet it has some drawbacks such aslower imaging speed and complex operation. In order to optimize this method, weimplement least-square algorithm into data processing to remove the component ofthe shaped function signal. This algorithm simplifies the imaging process and doublesthe imaging speed. Meanwhile, we apply a saw-tooth wave light as the shapedfunction signal to the image sensor by mounting a LED beside the camera lens. Thismodification makes the method adapted to the distant object imaging.(4) We design a low-light-level imaging system with our frame accumulation andshaped function technique. The hardware of the imaging system includes Altera’sDE2multimedia developmen board and a daughter card to generate the optical shapedfunction signal. The sofware of the imaging system contains the vedio signal decodemodule, vedio signal processing module, vedio signal DAC module and opticalshaped function signal DAC module. The low-light-level imaging system can get thebest fitting data matrix for the low-light-level image with least square algorithm.