Hu, BaoxinNaveed, Faizaan2019-11-222019-11-222019-052019-11-22http://hdl.handle.net/10315/36696In this study, an improved treetop detection and a region-based segmentation algorithm were developed to delineate Individual Tree Crowns (ITCs) using multispectral Light Detection and Ranging (LiDAR) data. The dataset used for this research was acquired from Teledyne Optechs Titan LiDAR sensor which was operated at three wavelengths: 1550 nm, 1064 nm, and 532 nm. An improved multi-scale method was developed to identify treetops for different crown sizes and merge them via Gaussian fitting. With the improved region growing segmentation method, neutrosophic logic was extensively used to incorporate contextual intensity information in the region merging decision heuristics. The LiDAR positional data was uniquely exploited, in this research, to generate refine crown boundary approximations. The results from the proposed method were compared with manually delineated ITCs to highlight the performance improvements. A 12% increase in the accuracy was observed with the proposed method over the popular Marker Controlled Watershed segmentation technique.Author owns copyright, except where explicitly noted. Please contact the author directly with licensing requests.Computer scienceElectronic Thesis or Dissertation2019-11-22Computer visionImage processingRemote sensingIndividual tree crown delineationRegion based segmentationScale analysisTreetop identificationLight detection and rangingLiDARGaussian residual analysisNeutrosophic logic