Integral Image Optimizations for Embedded Vision Applications

Kisačanin, Branislav

doi:10.1109/ssiai.2008.4512315

Cited by 23 publications

(10 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Kisačanin [22] describes three different recursions to achieve this and finds that the one suggested by Viola and Jones [35] has low memory requirements and low complexity. Denote the cumulative sum of the m-th row by R m,n .…”

Section: Mathematical Shortcutsmentioning

confidence: 99%

Algorithmic and software techniques for embedded vision on programmable processors

Kisačanin

Nikolic

2010

Signal Processing: Image Communication

View full text Add to dashboard Cite

Section: Mathematical Shortcutsmentioning

confidence: 99%

Algorithmic and software techniques for embedded vision on programmable processors

Kisačanin

Nikolic

2010

Signal Processing: Image Communication

View full text Add to dashboard Cite

“…The image integration in the original serial implementation is split into a dual-pass procedure where a prefix sum is computed on each row, followed by a prefix sum on each column of the output of the first pass-a common algorithm found in the literature, e.g., [23,24], for computing the integral image. Furthermore, to avoid extra cost of thread management, the computation in the first pass is assigned to the loading threads.…”

Section: Parallel Image Integrationmentioning

confidence: 99%

Computing Optimised Parallel Speeded-Up Robust Features (P-SURF) on Multi-Core Processors

Zhang

2009

Int J Parallel Prog

View full text Add to dashboard Cite

This article presents a novel CPU-based parallel algorithm (P-SURF) that computes the Speeded-Up Robust Features (SURF), a local descriptor that is able to find point correspondences between images in spite of scaling and rotation. The algorithm presented here parallelises all the seven major steps found in the original serial computation. The task in each of the steps is decomposed and the fractions are assigned to running threads bound onto distinctive processors. The implementation of the algorithm was tested using randomly selected images in regard to performance, scalability and stability. The results showed that its performance on mid-level Intel Core Duo processors was comparable to that of some fast GPU-based SURF implementations. For example, on a testing system equipped with an Intel Core Duo P8600 at 2.4 GHz, P-SURF was able to extract and represent features from a 640 × 480 image at a rate of 33 frames per second. The experimental results also revealed that, instead of leaving the threads to the kernel for processor assignment, assigning hard processor affinity by the algorithm produced better performance and stability.

show abstract

“…In a subsequent patent Kisacanin and Yoder [46] describe how the method of developing a classifier using adaboost-over-genetic programming is used to process real-time video image data in order to determine the open versus closed eye state of a human. They specifically mention how this process can be embedded in a microprocessor and used as part of a driver monitoring system.…”

Section: Gp For Classificationmentioning

confidence: 99%

Recent Patents on Genetic Programming

O’Neill¹,

Brabazon²

2009

CSENG

View full text Add to dashboard Cite

Genetic Programming is a form of Natural Computing which adopts principles from neo-Darwinian evolution to automatically solve problems. It is a model induction method in that both the structure and parameters of the solution are explored simultaneously. Genetic Programming is a particularly interesting method as it is claimed to be an invention machine, producing solutions to problems that are competitive and in some cases superior to those produced by human experts. Its best solutions have become patentable inventions in their own right. In this article we overview some of the recent patents relating to Genetic Programming over the past three years. In light of the number and diversity of patent applications during this period it is clear that Genetic Programming is a vibrant field of research, which is having a significant impact on real-world applications, and is demonstrating clear commercial potential.

show abstract

Integral Image Optimizations for Embedded Vision Applications

Cited by 23 publications

References 5 publications

Algorithmic and software techniques for embedded vision on programmable processors

Algorithmic and software techniques for embedded vision on programmable processors

Computing Optimised Parallel Speeded-Up Robust Features (P-SURF) on Multi-Core Processors

Recent Patents on Genetic Programming

Contact Info

Product

Resources

About