Home // ACHI 2022, The Fifteenth International Conference on Advances in Computer-Human Interactions // View article

Detection of Pesticide Mist Distribution to Avoid Spray Drift

Authors:
Chaitali Dutta
Oky Dicky Ardiansyah Prima
Kanayo Ogura
Koichi Matsuda
Shoichi Yuki

Keywords: pesticide-mist, semantic segmentation, convolution neural network, agriculture, computer vision

Abstract:
Agriculture has made great progress due to technological advances. Spraying pesticides plays an important role in protecting crops from insects and pests. Large mechanical sprayers have made it easier for farmers to spray large areas in a short time. However, there is a risk of pesticides being sprayed in unintended areas, causing damage to nearby fields and bodies of water. In this study, we propose an approach to detect pesticide mist distribution using the U-Net-based semantic segmentation technique. To train the semantic segmentation model, images of mist from sprinklers and gardening mist sprayers were used as training data. Our results show that the semantic segmentation technique could infer the distribution of pesticide mist. The extracted mist areas were found to exclude areas, such as workers, gardening poles, and clouds. Furthermore, we were able to estimate the Three-Dimensional (3D) distribution of mist over the field based on the mist distribution in the continuous frame images. For the current attempt, we did not define the density of mist in the training data, however, we would like to consider estimating the density of mist in the future.

Pages: 6 to 10

Copyright: Copyright (c) IARIA, 2022

Publication date: June 26, 2022

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-61208-982-9

Location: Porto, Portugal

Dates: from June 26, 2022 to June 30, 2022