Image recognition of Legacy blueberries in a Chilean smart farm through deep learning
Quiroza, Ignacio A.
Image recognition of Legacy blueberries in a Chilean smart farm through deep learning A. Quiroza, Ignacio ; H. Alférez, Germán
Agriculture is one of the most important pillars of development in Chile. However, it is expected that around the year 2030 there is going to be a gradual decrease in the number of farmers. Therefore, it is necessary to replace this workforce with technology and mechanization. One way to do this is through smart farms to leverage agricultural production. The contribution of this research work is a novel approach for deep-learning image recognition of Legacy blueberries in the rooting stage that can be used in smart farms in Chile. Legacy blueberry is a variety of Southern Highbush blueberry. This species constitutes 80% of the blueberry crops in Chile. Specifically, we propose an image recognition approach based on a convolutional neural network (CNN) to detect the presence of trays with living blueberry plants, the presence of trays without living plants, and the absence of trays. The average results of the evaluation of the predictive model are as follows: accuracy: 86%, precision: 86%, recall: 88%, and F1 score: 86%.
0168-1699
Convolutional neural networks
Smart farms
Image recognition
Legacy blueberry
Image recognition of Legacy blueberries in a Chilean smart farm through deep learning A. Quiroza, Ignacio ; H. Alférez, Germán
Agriculture is one of the most important pillars of development in Chile. However, it is expected that around the year 2030 there is going to be a gradual decrease in the number of farmers. Therefore, it is necessary to replace this workforce with technology and mechanization. One way to do this is through smart farms to leverage agricultural production. The contribution of this research work is a novel approach for deep-learning image recognition of Legacy blueberries in the rooting stage that can be used in smart farms in Chile. Legacy blueberry is a variety of Southern Highbush blueberry. This species constitutes 80% of the blueberry crops in Chile. Specifically, we propose an image recognition approach based on a convolutional neural network (CNN) to detect the presence of trays with living blueberry plants, the presence of trays without living plants, and the absence of trays. The average results of the evaluation of the predictive model are as follows: accuracy: 86%, precision: 86%, recall: 88%, and F1 score: 86%.
0168-1699
Convolutional neural networks
Smart farms
Image recognition
Legacy blueberry