2017, Vol. 2, Issue 2, Part A
A robust measure of pairwise distance estimation approach: RD-RANSACAuthor(s):
A method for detection of outliers is proposed which a Robust Distance to be used RANSAC (RD-RANSAC). A novel idea on hot to make RANSAC repeatable is presented, which will find the optimum set in nearly run for multi-model. Robust methods are capable of discriminating correspondence outliers, thus, obtaining better results. Our proposed method is an improvement of RANSAC which takes into account additional information of the quality of the matches to largely reduce the computational cost of the pair wise distance estimation by Rousseeuw’s Minimum Covariance Determinant (MCD). However, even in quite large samples, the chi-square approximation to the distance of the sample data from the MCD center with respect to the MCD shape is poor 
. RANSAC can only estimate one model for a particular data set. The two or more model exists; RANSAC may fail to find either one. The problem is hard as the number of outlier is usually large, possibly larger than 50%, thus powerful estimation techniques are need. Experiments with up to 80% outlier prove the efficiency of RANSAC 
. RANSAC is not always able to find the optimum set even for moderately contaminated sets and it usually performs badly when the number of inliers is less. However this work proposes a new robust method for pairwise distance estimation to combine the benefits of RANSAC algorithm, namely improved quality, reduced computational time and less parameter to adjust and powerful estimation techniques up to more than 80% outlier prove the efficiency.Pages: 31-34 | Views: 1264 | Downloads: 12Download Full Article: Click Here
How to cite this article:
Ravi J. A robust measure of pairwise distance estimation approach: RD-RANSAC. Int J Stat Appl Math 2017;2(2):31-34.