A convolutional neural network approach for counting and geolocating citrus-trees in UAV multispectral imagery
Visual inspection has been a common practice to determine the number of plants in orchards, which is a labor-intensive and time-consuming task. Deep learning algorithms have demonstrated great potential for counting plants on unmanned aerial vehicle (UAV)-borne sensor imagery. This paper presents a...
| Autores: | , , , , , , , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2020 |
| País: | Brasil |
| Institución: | Universidade Estadual Paulista (UNESP) |
| Repositorio: | Repositório Institucional da UNESP |
| Idioma: | inglés |
| OAI Identifier: | oai:repositorio.unesp.br:11449/197328 |
| Acceso en línea: | http://dx.doi.org/10.1016/j.isprsjprs.2019.12.010 http://hdl.handle.net/11449/197328 |
| Access Level: | acceso abierto |
| Palabra clave: | Deep learning Multispectral image UAV-borne sensor Object detection Citrus tree counting Orchard |
| Sumario: | Visual inspection has been a common practice to determine the number of plants in orchards, which is a labor-intensive and time-consuming task. Deep learning algorithms have demonstrated great potential for counting plants on unmanned aerial vehicle (UAV)-borne sensor imagery. This paper presents a convolutional neural network (CNN) approach to address the challenge of estimating the number of citrus trees in highly dense orchards from UAV multispectral images. The method estimates a dense map with the confidence that a plant occurs in each pixel. A flight was conducted over an orchard of Valencia-orange trees planted in linear fashion, using a multispectral camera with four bands in green, red, red-edge and near-infrared. The approach was assessed considering the individual bands and their combinations. A total of 37,353 trees were adopted in point feature to evaluate the method. A variation of a (0.5; 1.0 and 1.5) was used to generate different ground truth confidence maps. Different stages (T) were also used to refine the confidence map predicted. To evaluate the robustness of our method, we compared it with two state-of-the-art object detection CNN methods (Faster R-CNN and RetinaNet). The results show better performance with the combination of green, red and near-infrared bands, achieving a Mean Absolute Error (MAE), Mean Square Error (MSE), R-2 and Normalized Root-MeanSquared Error (NRMSE) of 2.28, 9.82, 0.96 and 0.05, respectively. This band combination, when adopting sigma = 1 and a stage (T = 8), resulted in an R-2, MAE, Precision, Recall and F1 of 0.97, 2.05, 0.95, 0.96 and 0.95, respectively. Our method outperforms significantly object detection methods for counting and geolocation. It was concluded that our CNN approach developed to estimate the number and geolocation of citrus trees in highdensity orchards is satisfactory and is an effective strategy to replace the traditional visual inspection method to determine the number of plants in orchards trees. |
|---|