Spatiotemporal Agriculture Modeling

Distributed Autonomous System Laboratory

Collaborators and Students: 

Dr. Hossein Mohamadipanah (postdoctoral associate at DAS Lab)

Dr. Chinmay Soman (UIUC, NRES)

Kevin Wolz (UIUC, Plant Biology)

Dr. Evan Delucia (UIUC, iSEE)

Dr. Angela Kent (UIUC, NRES)

The Woody Perennial Polyculture Experimental Site at UIUC

Project Description: 

The goal of this project is to develop cyber-tools to enhance sensing, modeling, and managing complex ecosystems that are the key to making agriculture truly sustainable by leveraging the beneficial role of biodiversity in ecosystems. Accordingly, in the DASLAB the HNK algorithm is developed to model growth of plants on the Polycultural fields, which consists of planting different types of plants in parallel lanes. The HNK algorithm demonstrated a great reduction in processing time over existing kernel methods while enjoying better approximation accuracies for high spatiotemporal scale dynamical systems.

Figure 1 : Recorded picture from the Polyculture field. 

 

Figure 2 : Extraction information from the segmented image by EM method.

 

Figure 3 Reconstruction of the growth of plants 

 

Figure 4 Comparison of the HNK method in terms of computational time and maximum error with the litrature 


Time-lapse of a field

Video: Time-Lapse video of a field