International Journal of Science, Engineering and TechnologyPlant Phenotyping Using Convolution Neural Networks
Authors- Prasanth Yadla
Abstract-Plant phenotyping of a plant is describing the visual or observable characteristics such as height, biomass, leaf shape and so on. In this project we are using phe- notyping to find the water requirement of the plant [1–3].The number of leaf tips, width of the plant and collars of the plant can be phenotypic indicators of water stress. Our aim is to apply Image Processing and Neural networks to given images to detect leaves and col- lars as key points. This would help in identifying the phenotype features of the plant in an automated fashion. Specifically, we used the Tensorflow Object Detection API, an open source framework built on top of Tensorflow to localize and identify leaf tips and collars. We have achieved over 55% accuracy in detecting the leaf tips using Inception Network out of the total manual count and also detected collars to a reasonable accuracy.
DOI: /10.61463/ijset.vol.12.issue6.371
