RECONSTRUCTION OF TWO-DIMENSIONAL OBJECTS USING RANSAC MODIFICATION

Abstract

   Abstract—RANSAC (Random Sample Consensus) can be used as mathematical model parameters of a set of observed data contains inlier and outlier data. In manufacture can also be used to build exact line equations from a number of data, by validating the data by looking at the inlier and outlier data of a project being worked on. To create a program in making this final project, the author uses two-dimensional data that is processed by LiDAR (Light Detection and Ranging), the author conducts data collection and research on the performance of two-dimensional object reconstruction using RANSAC modifications in Python. In the first experiment using data from LiDAR detection, the authors limit that the variable value of the residual threshold is above 3000, so it can be expressed as an imperfect image, from the first test it is produced with 92.85% perfect data, while for the second test it gets 52% data. Perfect. Reconstruction of 2D objects from LiDAR data can be validated using the RANSAC method. Ransac method is used to build the exact equation of the line from a number of data. So that the generated line can be a reference whether the amount of data to be observed is a correct 2-dimensional object reconstruction. How to validate by looking at the suitability of the inlier and outlier data from the RANSAC line equation. Inliers data will indicate whether the pattern of the reconstruction data is appropriate.