Researchers from Osaka University have managed the reconstruction of plant branch structures, focusing on points of interest like branch structures under leaves. This has been achieved using advanced image analysis together with artificial intelligence technology. This represents one of the first applications of this type of technology to botany.
The stud is of particular importance to fruit-bearing trees. By gaining insights into the growth of branches and leaves of individual trees, those whose livelihoods are based on the cultivation of fruit trees can learn about new methods for managing trees. This can help with maximizing fruit growth and with helping to maintain and protect the trees. Looking after fruit trees requires specialists and daily management, with most practices based on observation.
In larger orchards, fruit tree growers have begun to use cameras systems. What the new research does is add automatic 3D modeling of plant shapes and branch structures to the mix. The idea is to save time and to find out new information.
The Japanese researchers used the collected imaged to form three-dimensional reconstructions, from multiple images of varying viewpoints. Because the images could not see the hidden portions of plants, such as branch structures hidden under their leaves, artificial intelligence was developed to construct these hidden elements.
The following video shows the research in action:
The reconstructions were achieved by linking together the original image in a Bayesian deep learning framework to develop the three-dimensional reconstruction. This allowed for the existence probability of branches that are hidden under leaves.
It is hoped the research will assist in realizing future cultivation technology to enabled detailed daily management of branch and leaf level processes of growing plants to take place, especially in assisting where cultivators cannot see all aspects of the tree. It is also hoped the research will provide advice about improved trimming and pruning methods. Moreover, the development of the technology could assist with forecasting the future growth of plants.
The results of the research are set to be unveiled at the EEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2018), which takes place between June 18 and June 22, 2018. The research will later be published in the journal Computer Vision Foundation Open Access.
