Direct and accurate feature extraction from 3D point clouds of plants using RANSAC

While point clouds hold promise for measuring the geometrical features of 3D objects, their application to plants remains problematic. Plants are three dimensional (3D) organisms whose morphology is complex, varies from one individual another and changes over time. Objective measurement attributes in cloud domain increasingly attractive as techniques improve accuracy reduce computational Analysis data, however, not straightforward, due its discrete nature, imaging noise cluttered background. In this paper, we introduce a robust method direct analysis data. To end, generalise random sample consensus (RANSAC) algorithm data then use it model different plant organs. Since obtained multi-view stereo images, they often contaminated with considerable level noise, distortions out-of-distribution points. Key our approach RANSAC on cloud, making technique more undesirable outliers. We tested proposed Brassica grapevine by comparing estimated measurements extracted models manual ones taken actual plants. Our achieved R2>0.90 measured diameters branches stems while yielded R2>0.91 leaf angles branch Brassica. all cases, produced stable performance under background conventional methods failed work.