Articles
Flower detection using computer vision algorithm and three-dimensional mapping for automatic pollination using robotic platform
Article number
1433_31
Pages
243 – 250
Language
English
Abstract
Pollinators such as bees, butterflies, birds, and insects transfer pollen from one flower to another aiming the fertilization of plants.
This process is essential for the reproduction of many flowering plants.
Robot pollinators can supplement natural pollination by assisting in situations where natural pollinators may be insufficient or unavailable.
This could be particularly valuable in areas facing pollinator declines due to factors such as habitat loss, pesticide use or indoor farming.
In indoor farming systems such as vertical farms or greenhouses, where natural pollinators may not have access or may be impractical to introduce, robot pollinators offer a viable alternative for pollination.
This enables the cultivation of a wider range of crops in controlled environments, promoting agricultural diversification and year-round production.
In this paper an algorithm is presented to recognize flowers and predict 3d position of flowers.
The pollinator robot will use the algorithm to execute pollination operation.
A proof-of-concept robot pollinator is programmed to deliver precise and targeted pollination to specific plant species or individual flowers, optimizing pollination efficiency and reducing waste of resources such as pollen and energy.
The algorithm uses deep learning approach in order to allow training different plants.
Robot pollinators can be complementary but can work alongside existing pollinator populations.
This process is essential for the reproduction of many flowering plants.
Robot pollinators can supplement natural pollination by assisting in situations where natural pollinators may be insufficient or unavailable.
This could be particularly valuable in areas facing pollinator declines due to factors such as habitat loss, pesticide use or indoor farming.
In indoor farming systems such as vertical farms or greenhouses, where natural pollinators may not have access or may be impractical to introduce, robot pollinators offer a viable alternative for pollination.
This enables the cultivation of a wider range of crops in controlled environments, promoting agricultural diversification and year-round production.
In this paper an algorithm is presented to recognize flowers and predict 3d position of flowers.
The pollinator robot will use the algorithm to execute pollination operation.
A proof-of-concept robot pollinator is programmed to deliver precise and targeted pollination to specific plant species or individual flowers, optimizing pollination efficiency and reducing waste of resources such as pollen and energy.
The algorithm uses deep learning approach in order to allow training different plants.
Robot pollinators can be complementary but can work alongside existing pollinator populations.
Authors
F.B. Marin, L.D. Buruiană, M.G. Matache, G. Gurău, M. Marin
Keywords
detecting, images, pollination, computer vision, convolutional neural networks
Online Articles (41)