Deep Multi-Scale Feature Learning for Defocus Blur Estimation
This work addresses the problem of accurate blur estimation for computer vision applications, but it is incremental as it builds on existing edge-based and deep learning approaches.
The paper tackles defocus blur estimation from a single image by classifying edges and estimating blur only at pattern edges, then using guided filters for interpolation to produce a dense blur map with clear boundaries. It outperforms state-of-the-art methods in experiments on naturally defocused images, achieving a good balance between running time and accuracy.
This paper presents an edge-based defocus blur estimation method from a single defocused image. We first distinguish edges that lie at depth discontinuities (called depth edges, for which the blur estimate is ambiguous) from edges that lie at approximately constant depth regions (called pattern edges, for which the blur estimate is well-defined). Then, we estimate the defocus blur amount at pattern edges only, and explore an interpolation scheme based on guided filters that prevents data propagation across the detected depth edges to obtain a dense blur map with well-defined object boundaries. Both tasks (edge classification and blur estimation) are performed by deep convolutional neural networks (CNNs) that share weights to learn meaningful local features from multi-scale patches centered at edge locations. Experiments on naturally defocused images show that the proposed method presents qualitative and quantitative results that outperform state-of-the-art (SOTA) methods, with a good compromise between running time and accuracy.