CMOS影像感測器色偏影像之研究

Abstract

近年來CMOS(Complementary Metal Oxide Semiconductor)影像感測器逐漸取代其競爭對手CCD(Charge Coupled Device)影像感測器，成為影像感測器市場主流是無庸置疑的事實，它被廣泛地應用在眾多領域，例如：智慧型手機、平板電腦，行車紀錄器，防盜安全監控裝置，甚至將來智慧車所需之大量影像感測器裝置...等，在在顯示其可預見未來之龐大市場與商機。 本論文以色偏影像之CMOS影像感測器樣本進行研究，首先從影像分析與故障分析著手，找出導致異常影像之原因與實體的缺陷所在，接著利用FDTD(Finite Difference Time Domain)光學軟體去模擬不同畫素結構訊號強度，最後再與CP(Circuit Probing)測試所獲得之實際影像訊號相比較，藉以驗證用FDTD光學模擬去評估CMOS影像感測器之單位畫素的量子效應與影像訊號強度之準確性。

Without a doubt, CMOS(Complementary Metal-Oxide-Semiconductor) image sensor, replacing CCD(Charge-Coupled Device) image sensor, is the mainstream of image sensor market in recent years. CMOS image sensor is used in many kinds of application, such as smart phone, tablet, vehicle video record, surveillance camera, and the visual devices in future smart car. So it has huge market and infinite business. This thesis is base on CMOS image sensor sample with discolor phenomenon for study. First, we use image and failure analysis to find out the root cause of discolor image as well as the physical defect location, then simulate image signal with different pixel structure by FDTD(Finite Difference Time Domain) optical simulation software. Finally, we compare the simulation result and actual image signal provided by CP(Circuit Probing) testing then to verify the accuracy of FDTD method for the estimation of quantum efficiency in unit pixel and image signal of CMOS image sensor.