Stereo matching method for fast high-resolution images effective large scale stereo matching

"Effective large scale stereo matching"

In this paper we propose a novel approach to binocular stereo for fast matching of high-resolution images. Our approach builds a prior in the disparities by forming a triangulation on a set of support points which can be robustly Matched, reducing the matching ambiguities of the remaining points. This allows for efficient exploitation of the disparity search space, yielding accurate dense reconstruction without the N Eed for global optimization.

The article published in the 2010 ACCV above, look at the formula and code are very complex, the main idea is to match the feature points on the image as a matching support point, to do triangulation on the support point, the difference in the interpolation calculation, but the effect is general, applicable to the occasion of high real-time requirements.

The whole project is done with OpenCV and OpenMP, the code is very complicated, but it fits the code style of the engineer, can learn, but the improvement is more difficult.

Effect: Obviously see the effect and the general global optimization algorithm still have a certain gap.

Article download Link:

http://www.cvlibs.NET/software/libelas/

Http://www.cvlibs.net/publications/Geiger2010ACCV.pdf

Code DOWNLOAD Link:

Http://www.cvlibs.net/download.PHP?file=libelas.zip

OpenMP version: Very short run time about 0.1s completely meet the real-time requirements.

Http://www.cvlibs.net/download.php?file=libelas_omp.zip

I am commissioning the Good project:

http://download.csdn.net/detail/wangyaninglm/9321831