AgriField3D
收藏资源简介:
AgriField3D是由爱荷华州立大学提供的一个专门针对玉米的三维点云和程序化模型数据集,旨在为人工智能在农业研究中的应用提供支持。该数据集包含了超过1000个高质量的三维点云,通过地面激光扫描技术捕获了田间生长的玉米植株。数据集还包含了520个经过语义和实例分割的植株模型,以及与实际植株数据相对应的程序化生成的玉米叶片模型。这些模型利用非均匀有理B样条(NURBS)进行表面表示,并通过粒子群优化和可微编程进行优化。数据集提供了多种分辨率版本的子采样,以满足不同的计算需求,并包含了详细的植株形态和质量元数据。
AgriField3D is a 3D point cloud and procedural model dataset dedicated to maize, provided by Iowa State University, with the aim of supporting the application of artificial intelligence in agricultural research. This dataset contains over 1,000 high-quality 3D point clouds of field-grown maize plants captured via terrestrial laser scanning. It also includes 520 plant models annotated for semantic and instance segmentation, as well as procedurally generated maize leaf models corresponding to real plant data. These models adopt Non-Uniform Rational B-Splines (NURBS) for surface representation and are optimized through particle swarm optimization and differentiable programming. The dataset offers subsampled versions with multiple resolutions to meet diverse computational demands, and contains detailed plant morphology and quality metadata.




