Occlusion and Non-occlusion edge database

Please click here to download ‘Extracted edge database.zip’ for the database of the extracted occlusion and non-occlusion edges. The zip file when extracted contains two folders. The contains of each folder is listed below

Folder: Hand-traced OcclusionEdges

This folder contains the extracted occlusion edges which were first hand-traced by subjects on high resolution images from the Mcgill Color Image Database. The folder ‘Hand-traced OcclusionEdges’ contains 38 sub-folders for 38 selected natural images. Each sub folder contains a high-resolution image in Tiff format, hand-traced edge image (‘ed_1.bmp’) and a folder ‘80x40’ which contains the extracted edge patches. The naming convention for the extracted edge patches is as follows:

sub_1_edge_@(172,1687).mat : This patch is centered at row 172 and column 1687.

Folder: Canny detected Occlusion and Non-occlusion edges

This folder contains the extracted occlusion and non-occlusion edges which were first detected by Canny algorithm. The naming convention for the extracted edge patches is as follows:

img_1_edge_@(372,776).mat : This patch is extracted from the image contained in Sub-folder 1 of the folder ‘Hand-traced OcclusionEdges’. The patch is centered at row 372 and column 776.

If you use our database, please cite following two articles:

  • Vilankar, K. P., Golden, J. R., Chandler, D. M., & Field, D. J. (2014). Local edge statistics provide information regarding occlusion and nonocclusion edges in natural scenes.Journal of vision,14(9),13.
  • Olmos, A., Kingdom, F. A. A. (2004). A biologically inspired algorithm for the recovery of shading and reflectance images.Perception,33(12), 1463-1473.