A Fast Approximation of the Bilateral Filter Using a Signal Processing Approach

By Sylvain Paris, Frédo Durand. International Journal of Computer Vision, 2006

1,190 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/2c8cbeb547f3c982c820794ec9cf431b133aba57

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Normalized Cut Loss for Weakly-Supervised CNN Segmentation (2018, methodology): “Gaussian kernel for 5-d RGBXY, this comes down to solving a bilateral filtering with many available fast computation techniques [37, 2, 1].”
Spectral–Spatial Hyperspectral Image Classification With Edge-Preserving Filtering (2014, methodology): “2) Analysis of the Influence of Different Postprocessing Techniques: For the proposed spectral–spatial classification framework, different postprocessing techniques, e.g., other types of edge-preserving smoothing methods or statistics-based method, can be used to refine the binary probability maps.”
Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials (2011, methodology): “This implies that approximate message passing can be performed in O(N) time [16].”
Color Balance and Fusion for Underwater Image Enhancement (2018, background): “A common solution to overcome this limitation is to employ multi-scale linear [57], [59] or non-linear filters [60], [61].”
Learning Sparse High Dimensional Filters: Image Filtering, Dense CRFs and Bilateral Neural Networks (2015, background): “(1) have been developed, including the bilateral grid [36], Gaussian KDtrees [3], and the permutohedral lattice [2].”
Joint bilateral upsampling (2007, methodology): “Durand and Dorsey [2002] applied the bilateral filter to HDR tone mapping and also described a fast approximation, which was recently improved upon [Paris and Durand 2006; Weiss 2006].”
A non-local cost aggregation method for stereo matching (2012, methodology): “A number of approximation methods have been developed to accelerate the bilateral filter, including Paris and Durand’s fast bilateral filter [12], Porikli’s O(1) bilateral filter [17] and Yang’s real-time bilateral filters [25, 26].”
Image smoothing via L0 gradient minimization (2011, methodology): “Our method relates in spirit to edge-preserving smoothing [Tomasi and Manduchi 1998; Durand and Dorsey 2002; Paris and Durand 2006; Farbman et al. 2008; Subr et al. 2009; Kass and Solomon 2010] that aims to retain primary color change, and yet differs from them in essence in focus and in mechanism.”
Fast Global Image Smoothing Based on Weighted Least Squares (2014, background): “The base layer captures the main structural information, while the detail layer contains the residual smaller scale details in the image.”
Nonlinear disparity mapping for stereoscopic 3D (2010, methodology): “This isotropic distance measure in image and disparity space results in a locally adaptive anisotropic filter in image space only (similar to the idea of the Fast Bilateral Filter [Paris and Durand 2006]).”
Fast Image Processing with Fully-Convolutional Networks (2017, background): “The bilateral filter in particular has benefitted from long-term investment in its acceleration [21, 72, 15, 59, 2, 1, 29, 8].”
Disparity map quality for image-based rendering based on multiple metrics (2014, neutral): “filter with σ = 11), as well as a median filter (‘x’, 51 × 51 window) and a cross-bilateral grid filter (‘*’, as described by Paris and Durand [8] and implemented in Matlab by Jiawen Chen).”
Urban-Area and Building Detection Using SIFT Keypoints and Graph Theory (2009, background): “There is also a faster version of BF proposed by Paris and Durand [24].”
Multiresolution Bilateral Filtering for Image Denoising (2008, neutral): “[33] later derives an improved acceleration scheme for the filter through expressing it in a higher dimensional space where the signal intensity is added as the third dimension to the spatial domain.”
Real-time O(1) bilateral filtering (2009, background): “Several efficient numerical schemes [9, 17, 18, 19, 20] enable it to be computed at”
Edge-avoiding wavelets and their applications (2009, background): “…N since subsampling is avoided [Fattal et al. 2007], solving multiple linear systems [Farbman et al. 2008], coping with the resulting poorly-conditioned systems [Szeliski 2006], or introducing additional dimensions and additional memory cost due to their discretization [Paris and Durand 2006].”
High Dynamic Range Imaging (2015, neutral): “Since the 2002 publication, several faster algorithms for computing bilateral filter have been proposed [1, 2, 23, 113].”