US 12,148,197 B2
System and method for plant disease detection support
Artzai Picon Ruiz, Derio (ES); Matthias Nachtmann, Limburgerhof (DE); Maximilian Seitz, Limburgerhof (DE); Patrick Mohnke, Limburgerhof (DE); Ramon Navarra-Mestre, Limburgerhof (DE); Alexander Johannes, Limburgerhof (DE); Till Eggers, Ludwigshafen (DE); Amaia Maria Ortiz Barredo, Vitoria-Gasteiz (ES); Aitor Alvarez-Gila, Derio (ES); and Jone Echazarra Huguet, Derio (ES)
Assigned to BASF SE, Ludwigshafen am Rhein (DE)
Appl. No. 17/611,517
Filed by BASF SE, Ludwigshafen am Rein (DE)
PCT Filed May 14, 2020, PCT No. PCT/EP2020/063428
§ 371(c)(1), (2) Date Nov. 15, 2021,
PCT Pub. No. WO2020/229585, PCT Pub. Date Nov. 19, 2020.
Claims priority of application No. 19174907 (EP), filed on May 16, 2019.
Prior Publication US 2022/0230305 A1, Jul. 21, 2022
Int. Cl. G06V 10/764 (2022.01); G06N 3/048 (2023.01); G06T 7/00 (2017.01); G06T 7/10 (2017.01); G06V 10/82 (2022.01); G06V 20/10 (2022.01)
CPC G06V 10/764 (2022.01) [G06N 3/048 (2023.01); G06T 7/0012 (2013.01); G06T 7/10 (2017.01); G06V 10/82 (2022.01); G06V 20/188 (2022.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30188 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A computer-implemented method (1000) for detecting plant diseases using a convolutional neural network (120) trained with a multi-crop dataset, the training dataset comprising training input images showing various crops, and each of the training input images shows a part of a particular crop with one or more disease symptoms associated with one or more diseases of interest, or shows a part of a particular crop with abiotic marks, or shows a healthy part of a particular crop, the convolutional neural network (120) with an extended topology comprising an image branch (121) based on a classification convolutional neural network for classifying test input images according to plant disease specific features, a crop identification branch (122) for adding plant species information, and a branch integrator for integrating the plant species information with each test input image, the plant species information (20) specifying the crop on the respective test input image (10), the method comprising:
receiving (1100) a test input comprising an image of a particular crop showing one or more particular plant disease symptoms, and a respective crop identifier;
applying (1200) the trained convolutional neural network (120) to the received test input; and
providing (1300) a classification result according to the output vector of the convolutional neural network, the classification result indicating the one or more plant diseases associated with the one or more particular plant disease symptoms.