Three-Dimensional Separable Steerable Filters

Contributors

• Konstantinos G. Derpanis
• Jacob M. Gryn
• Markus Enzweiler

Abstract:

This work details the construction of three-dimensional (e.g., X-Y-Z) separable steerable filters. The approach presented is an extension of the construction of two-dimensional (X-Y) separable steerable filters. Additionally, three-dimensional separable steerable filters, both continuous and discrete versions, for the second derivative of the Gaussian and its Hilbert transform are reported. Experimental evaluation demonstrate that the error in the constructed separable filters are negligible. Finally, an application of the filters for surveillance based on visual motion is presented.

Related Publications:

• K.G. Derpanis and J.M. Gryn, Three-Dimensional nth Derivative of Gaussian Separable Steerable Filters, ICIP 2005, vol III, pp553-556
• K.G. Derpanis and J.M. Gryn, Three-Dimensional nth Derivative of Gaussian Separable Steerable Filters, York University Technical Report CS-2004-05, November 22, 2004

Related Code and Information:

• ImgProc C++ Library (Jacob Gryn - rewrite and expansion of original 3DSteer)
• Original 3DSteer C Library (Markus Enzweiler)
• 3DSteer Matlab Library (Konstantinos Derpanis) On this site
• PVN File Format Specifications and C code (Jacob Gryn - needed for ImgProc)