Methodology for Plant Specific Cultivation through a Plant Identification pipeline

Agriculture needs to optimization for satisfying the rising demands of food due to world population growth. An approach to this problem is digitization of agriculture through IT tools by creating digital twins. However, in the digital twin creation the uniqueness of the plant is lost, therefore, plant based agricultural cultivation cannot be performed. Hence, this paper proposes a methodology to assign an identification marker to plants in a crop using an image analysis pipeline. To show the effectiveness of the algorithm the proposed method is evaluated on the Rovitis robotic platform and compared with the crop ontology. The outcome of this work can be used in robotic agricultural platforms to address plants singularly thus optimizing their cultivation.