Problem
Research questions and friction points this paper is trying to address.
Shape from Focus
focal stack
depth reconstruction
minimal sampling
practical applicability
Towards Minimal Focal Stack in Shape from Focus
🤖 AI Summary
This work proposes an efficient depth estimation framework that requires only two input images, addressing the practical limitations of traditional Shape from Focus (SFF) methods which rely on densely sampled focal stacks. The approach first synthesizes an all-in-focus image and an energy difference map through a physics-driven focal stack augmentation strategy, then iteratively refines the depth estimate using a multi-scale ConvGRU network. To the best of our knowledge, this is the first method to achieve high-precision depth reconstruction from just two images, substantially reducing data acquisition costs while maintaining state-of-the-art performance. Extensive experiments on both synthetic and real-world datasets demonstrate the effectiveness and robustness of the proposed framework.
📝 Abstract
Shape from Focus (SFF) is a depth reconstruction technique that estimates scene structure from focus variations observed across a focal stack, that is, a sequence of images captured at different focus settings. A key limitation of SFF methods is their reliance on densely sampled, large focal stacks, which limits their practical applicability. In this study, we propose a focal stack augmentation that enables SFF methods to estimate depth using a reduced stack of just two images, without sacrificing precision. We introduce a simple yet effective physics-based focal stack augmentation that enriches the stack with two auxiliary cues: an all-in-focus (AiF) image estimated from two input images, and Energy-of-Difference (EOD) maps, computed as the energy of differences between the AiF and input images. Furthermore, we propose a deep network that computes a deep focus volume from the augmented focal stacks and iteratively refines depth using convolutional Gated Recurrent Units (ConvGRUs) at multiple scales. Extensive experiments on both synthetic and real-world datasets demonstrate that the proposed augmentation benefits existing state-of-the-art SFF models, enabling them to achieve comparable accuracy. The results also show that our approach maintains state-of-the-art performance with a minimal stack size.
Innovation
Methods, ideas, or system contributions that make the work stand out.
focal stack augmentation
Shape from Focus
all-in-focus image
Energy-of-Difference
ConvGRU
🔎 Similar Papers
No similar papers found.
