Plant Point Cloud

Plant point clouds represent 3D models of plants generated from various sensing technologies, primarily aiming to automate the extraction of plant phenotypic traits for applications in agriculture and botany. Current research focuses on developing and evaluating deep learning methods, including neural radiance fields (NeRFs) and transformer-based architectures, for accurate semantic and instance segmentation of these point clouds, often addressing challenges posed by complex plant structures and limited annotated data. These advancements enable efficient and precise 3D plant modeling, facilitating high-throughput phenotyping and improving plant breeding and crop management strategies.