Digital imaging system testing and design using physical sensor characteristics

Image sensor testing and image quality enhancement methods that are geared towards commercial CMOS image sensors are developed in this thesis. The methods utilize sensor characterization data and camera system image processing information in order to improve their performance.Photo Response Non-Uniformity. An image sensor system-level pixel-to-pixel photo-response non-uniformity (PRNU) error tolerance method is presented in Chapter 2. A novel scheme is developed to determine sensor PRNU acceptability and corresponding sensor application categorization. Excessive variation in the sensitivity of pixels is a significant cause of the screening rejection for low-cost CMOS image sensors. The proposed testing methods use the concept of acceptable degradation applied to the camera system processed and decoded images. The analysis techniques developed give an estimation of the impact of the sensor's PRNU on image quality. This provides the ability to classify the sensors for different applications based upon their PRNU distortion and error rates.Perceptual criteria are used in the determination of acceptable sensor PRNU limits. These PRNU thresholds are a function of the camera system's image processing and sensor noise sources. We use a Monte Carlo simulation solution and a probability model-based simulation solution along with the sensor models to determine PRNU error rates and significances for a range of sensor operating conditions. We develop correlations between conventional industry PRNU measurements and final processed and decoded image quality thresholds. The results show that the proposed PRNU testing method can reduce the rejection rate of CMOS sensors.Cross-Talk Correction. A simple multi-channel imager restoration method utilizing a priori sensor characterization information is presented in Chapter 3. A novel method is developed to correct the channel dependent cross-talk of a Bayer color filter array sensor with signal-dependent additive noise. We develop separate cost functions (weakened optimization) for each color channel component-to-color channel component. Regularization is applied to each color channel component-to-color channel component, instead of the standard per color channel basis (giving us four optimal regularization parameters per color channel). This separation of color components allows us to calculate regularization parameters that take advantage of the differing magnitudes of each color channel component-to-color channel component cross-talk blurring, resulting in an improved trade-off between inverse filtering and noise smoothing.The restoration solution has its regularization parameters determined by maximizing the developed local pixel SNR estimations. The restoration method is developed with the goal of viable implementation into the on-chip digital logic of a low-cost CMOS sensor. The separate color channel component-to-color channel component approach simplifies the problem by allowing a set of four independent color channel component optimizations per pixel. Local pixel adaptivity can also be easily applied. Performance data of the proposed correction method is presented using color images captured from low cost embedded imaging CMOS sensors.