A New Robust Estimator with Application to Image Analysis

The field of computer vision is undergoing tremendous development in recent years. Computer vision concerns with developing systems that can interpret the content of natural scenes. Robust statistical methods were first adopted in computer vision to improve the performance of feature extraction algorithms at the bottom level of the vision hierarchy. These methods tolerate the presence of data points that do not obey the assumed model such points are typically called “outlier”. Recently, various robust statistical methods have been developed and applied to computer vision tasks. Random Sample Consensus (RANSAC) estimators are one of the widely applied to tackle such problems due to its simple implementation and robustness. In this paper we propose a novel and highly robust estimator, called INAPSAC (Improved N Adjacent Points Sample Consensus). The performance of the proposed algorithm has been studied through the experimental study along with the existing algorithms in the context of RANSAC techniques under the characteristics such as number of corners detected in an image with various types and the processing time is also considered.