GPU-efficient recursive filtering and summed-area tables

Abstract

Image processing operations like blurring, inverse convolution, and summed-area tables are often computed efficiently as a sequence of 1D recursive filters. While much research has explored parallel recursive filtering, prior techniques do not optimize across the entire filter sequence. Typically, a separate filter (or often a causal-anticausal filter pair) is required in each dimension. Computing these filter passes independently results in significant traffic to global memory, creating a bottleneck in GPU systems. We present a new algorithmic framework for parallel evaluation. It partitions the image into 2D blocks, with a small band of additional data buffered along each block perimeter. We show that these perimeter bands are sufficient to accumulate the effects of the successive filters. A remarkable result is that the image data is read only twice and written just once, independent of image size, and thus total memory bandwidth is reduced even compared to the traditional serial algorithm. We demonstrate significant speedups in GPU computation.

A Tutorial on the Implementations of Linear Image Filters in CPU and GPU

“…Finally, if we need a multiscale version of the box filter, the use of integral images is a good solution. In [5,2] the authors compare GPU and CPU implementation of integral images.…”

A Fresh Look at Generalized Sampling

“…A modern parallelization of recursive filters for GPUs recently achieved over 6GiPix/s on an NVIDIA GTX 480 GPU for the bicubic B-spline interpolation problem [Nehab et al, 2011]. The key message is that, whatever the approach followed or architecture used, the digital processing stage in generalized sampling is extremely fast.…”

Two-way partitioning of a recursive Gaussian filter in CUDA

“…Thus, its parallelism is restricted by the height or width of the input image. Other papers have partitioned image domains in a similar manner while implementing similar algorithms [17,21,22].…”