Coded Pattern Optimization: Image Model-based Theoretical Analysis

Coded exposure photography (CEP) modulates a point spread function (PSF) causing motion blur to enhance the motion deblurring performance. The performance of motion deblurring using CEP depends on the coded pattern which modulates the PSF, thus, the optimality of the coded pattern is the key to successful deblurring. Since the intensity of the captured image has a non-linear relation with the irradiance of the scene, the CEP without considering the non-linear relationship in deblurring is very likely to have poor deblurring results. In this paper, we investigate how to optimize the pattern by considering the nonlinear characteristics of image capturing.