Mapillary based plant distributions of ethnobotanical afforestation.

Abstract

Abstract: Mapillary is an open-source code base for the use of GPU based Deep Learning for Semantic Segmentation of wild images. We propose the creation of an autonomous drone for the automated capture of scientific images of medicinal and edible plants to create geotagged maps of plants on Mapillary.com with additional tags on plant sizes, species, and edible and medicinal value. This information is used in the planning of sponsored five or more level afforestation as social and academic forestry for edible and medicinal value. The same research is also useful in planning afforestation on Mars. Keywords: Miyawakis, Mapillary, Seamless Segmentation, FPN, ResNet50, Redtail, Edible and Medicinal Plants, Geotag