VisND: A Visualization Tool for Multidimensional Model of Canopy
| Autor: | Ali Shafiekhani, Guilherme N. DeSouza, Felix B. Fritschi |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Modality (human–computer interaction)
business.industry Computer science 010103 numerical & computational mathematics 02 engineering and technology Solid modeling Iterative reconstruction computer.software_genre 01 natural sciences Field (computer science) Visualization 0202 electrical engineering electronic engineering information engineering Data analysis 020201 artificial intelligence & image processing Artificial intelligence Data mining 0101 mathematics business computer Graphical user interface |
| Zdroj: | CVPR Workshops |
| DOI: | 10.1109/cvprw.2019.00324 |
| Popis: | Plant phenotyping is a data-driven research where interpretation of large and often multidimensional data is required. Therefore, effective visualization of plant phenotypes plays a major role in data analytics as it can provide scientists with the required tool to extract and infer important information. In that sense, unifying the large and multidimensional phenotypical data into one single model of the canopy can help plant biologists to correlate information from different dimensions and derive new observations and understandings in plant sciences. In this paper, we proposed a spatio-temporal tool for high-dimensional modeling and visualization of canopy for plant phenotyping. The goal is to offer an open-source visualization tool, named VisND (for N-Dimensional), that will provide a Graphical User Interface (GUI) where plant scientists can easily extract and analyze multidimensional models, registered over time, from different sensors and viewpoints. For this paper, we created 5D models (3D-RGB and Temperature over Time) of a crop by fusing and registering data captured using our field-based phenotyping platform: Vinoculer, a trinocular, multi-spectrum, observation tower. The platform is part of a study on the behavior of plants in response to different biotic and/or abiotic stresses. The data was captured using Infrared Thermography (IRT) along with multiview, visible imaging technology over an entire planting season and on a 24/7 basis. While currently VisND is being demonstrated for 5D models, it can be easily extended to incorporate other modality of sensors (more dimensions) and from other sources, such as other robotic platforms also operating in the field e.g. our mobile robot, Vinobot. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|