Useful Code



This page collects publicly-available resources and code that are useful for visual effects.

General Programs

• Irfanview
• GIMP
• OpenCV
• Paraview

Image Matting

• Alpha Matting Evaluation for benchmarking matting algorithms
• Closed-form matting code by A. Levin
• Spectral matting code by A. Levin
• GrabCut implementation by P. Wang

Image Compositing and Editing

• Matlab Laplacian pyramid toolbox by M. Do
• Photomontage by A. Agarwala et al.
• PDE-based inpainting by W. Yung and A.J. Ahankar
• Patch-based inpainting implementation by S. Bhat
• Saliency toolbox by D. Walther
• Optimized scale-and-stretch resizing by Y. Wang et al.
• Seam Carving by M. Rubinstein
• ImageStack by A. Adams et al.
• PatchMatch by C. Barnes et al.

Features and Matching

• Affine feature detectors and descriptors by K.Mikolajczyk
• Laplacian- and Difference-of-Gaussian detectors by W.-L. Zhao
• SIFT detector and descriptor by D. Lowe
• VLFeat (includes a Matlab SIFT implementation) by A. Vedaldi and B. Fulkerson
• QR Code generator by J. Delport
• ARToolkit markers by M. Billinghurst et al.
• ARTag markers by M. Fiala

Dense Correspondence and its Applications

• The Middlebury optical flow benchmark by S. Baker et al.
• Optical flow code by D. Sun et al.
• Human-assisted motion annotation by C. Liu et al.
• Fundamental matrix estimation and other Matlab functions for multiview geometry by P. Kovesi
• Image rectification by M. Dailey
• The Middlebury stereo benchmark by D. Scharstein and R. Szeliski
• Graph cut and belief propagation code by R. Szeliski et al.
• Max-flow/min-cut code by V. Kolmogorov
• Robust homography estimation and other Matlab functions for multiview geometry by D. Capel et al.

Matchmoving

• Camera calibration toolbox by J.-Y. Bouguet
• Blender

Motion Capture

• CMU motion capture database
• Biomechanical toolkit (including Mokka visualizer)
• Matlab motion capture toolbox by N. Lawrence
• Matlab visual hull code by B. Tordoff
• DAZ 3D has some useful free 3D models

3D Data Acquisition

• The Middlebury multiview stereo benchmark by S. Seitz et al.
• Patch-based multiview stereo (PMVS) by Y. Furukawa and J. Ponce
• Normal estimation and spin image code by A. Mian et al.
• Poisson surface reconstruction by M. Kazhdan