A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Deep Learning for Smart Plant Weed Applications Employing an Unmanned Aerial Vehicle
The investigation carried out in this paper elucidates the work to develop and test a smartweed detector and herbicide sprayer that utilizes a weed detection module for weed eradication. Through the years, weeds have remained a tremendous constant threat to the overall production of desired crops or farming productivity. Hence, an agile timely and accurate management of weeds could tremendously extenuate economic losses globally, denigrate an overuse of herbicides that sabotage the environment and revolutionize the agricultural sector. This paper further proposes an approach for emerging technology or recent advancement of deep learning by building a model through constructing and training a Convolutional Neural Network (CNN) that features real-time object detection on Raspberry Pi. Further details on this principle of operation are provided in this paper. A Convolutional Neural Network utilizing transfer learning was trained on the TensorFlow framework and yielded training and validation accuracies of 89.6% and 90.6% respectively. It was pre-trained using the weights from the Inception V3 architecture to detect multiple classes of weeds and crops. The sprayer module is further integrated to control sprayer operation, and it features an efficient chemical application.
Deep Learning for Smart Plant Weed Applications Employing an Unmanned Aerial Vehicle
The investigation carried out in this paper elucidates the work to develop and test a smartweed detector and herbicide sprayer that utilizes a weed detection module for weed eradication. Through the years, weeds have remained a tremendous constant threat to the overall production of desired crops or farming productivity. Hence, an agile timely and accurate management of weeds could tremendously extenuate economic losses globally, denigrate an overuse of herbicides that sabotage the environment and revolutionize the agricultural sector. This paper further proposes an approach for emerging technology or recent advancement of deep learning by building a model through constructing and training a Convolutional Neural Network (CNN) that features real-time object detection on Raspberry Pi. Further details on this principle of operation are provided in this paper. A Convolutional Neural Network utilizing transfer learning was trained on the TensorFlow framework and yielded training and validation accuracies of 89.6% and 90.6% respectively. It was pre-trained using the weights from the Inception V3 architecture to detect multiple classes of weeds and crops. The sprayer module is further integrated to control sprayer operation, and it features an efficient chemical application.
Deep Learning for Smart Plant Weed Applications Employing an Unmanned Aerial Vehicle
Ukaegbu, Uchechi (author) / Mathipa, Ledile (author) / Malapane, Maite (author) / Tartibu, Lagouge K. (author) / Olayode, Isaac O. (author)
2022-05-25
3872391 byte
Conference paper
Electronic Resource
English
Unmanned aerial vehicle airport and unmanned aerial vehicle system
| European Patent Office | 2020
| European Patent Office | 2024
Application of Deep Learning and Unmanned Aerial Vehicle on Building Maintenance
| DOAJ | 2021
Unmanned aerial vehicle base station and unmanned aerial vehicle system
| European Patent Office | 2024