Image and Depth from a Single Defocused Image Using Coded Aperture Photography
This work addresses challenges in computational photography for applications like robotics and imaging, but it is incremental as it builds on existing coded aperture techniques.
The paper tackles the problems of depth from defocus and defocus deblurring from a single image by proposing a multi-objective function to design an asymmetric coded aperture pattern, which improves results for both tasks, with extensive simulations and real experiments showing comparisons to previous methods.
Depth from defocus and defocus deblurring from a single image are two challenging problems that are derived from the finite depth of field in conventional cameras. Coded aperture imaging is one of the techniques that is used for improving the results of these two problems. Up to now, different methods have been proposed for improving the results of either defocus deblurring or depth estimation. In this paper, a multi-objective function is proposed for evaluating and designing aperture patterns with the aim of improving the results of both depth from defocus and defocus deblurring. Pattern evaluation is performed by considering the scene illumination condition and camera system specification. Based on the proposed criteria, a single asymmetric pattern is designed that is used for restoring a sharp image and a depth map from a single input. Since the designed pattern is asymmetric, defocus objects on the two sides of the focal plane can be distinguished. Depth estimation is performed by using a new algorithm, which is based on image quality assessment criteria and can distinguish between blurred objects lying in front or behind the focal plane. Extensive simulations as well as experiments on a variety of real scenes are conducted to compare our aperture with previously proposed ones.