This paper presents a method for automatic fruit detection in vineyards through the inspection of colour images obtained by a low cost RGB-D sensor placed on-board an agricultural vehicle. Image segmentation is obtained by using a pre-trained convolutional neural network, which receives input data, as sub-patches of known size, and performs the classification in a few classes of interest. Output scores are then used to create the probability maps for each class and, thus, pixel-by-pixel segmentation of the grape clusters. Field experiments prove the ability of the proposed processing to successfully segment grape clusters, with accuracy of 87.5%, despite the poor quality of the input images.
Deep learning-based image segmentation for grape bunch detection / Marani, R.; Milella, A.; Petitti, A.; Reina, G.. - STAMPA. - (2019), pp. 791-797. (Intervento presentato al convegno 12th European Conference on Precision Agriculture, ECPA 2019 tenutosi a Montpellier, France nel July 8-11, 2019) [10.3920/978-90-8686-888-9_98].
Deep learning-based image segmentation for grape bunch detection
Marani R.;Milella A.;Petitti A.;Reina G.Conceptualization
2019-01-01
Abstract
This paper presents a method for automatic fruit detection in vineyards through the inspection of colour images obtained by a low cost RGB-D sensor placed on-board an agricultural vehicle. Image segmentation is obtained by using a pre-trained convolutional neural network, which receives input data, as sub-patches of known size, and performs the classification in a few classes of interest. Output scores are then used to create the probability maps for each class and, thus, pixel-by-pixel segmentation of the grape clusters. Field experiments prove the ability of the proposed processing to successfully segment grape clusters, with accuracy of 87.5%, despite the poor quality of the input images.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
