Non-invasive detection of pesticide residues in freshly harvested olives using hyperspectral imaging technology
Pesticides play a crucial role in boosting the overall yield and productivity of agricultural produce by controlling pests, insects, and various plant diseases. However, excessive use of pesticides has led to contamination of food products and water bodies, as well as disruption of ecological and en...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión aceptada para publicación |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universidad de Jaén |
| Repositorio: | RUJA. Repositorio Institucional de la Producción Científica de la Universidad de Jaén |
| OAI Identifier: | oai:ruja.ujaen.es:10953/4144 |
| Acceso en línea: | https://doi.org/10.1016/j.atech.2024.100644 https://hdl.handle.net/10953/4144 |
| Access Level: | acceso abierto |
| Palabra clave: | Freshly harvested olives Computer vision Image processing Hyperspectral imaging Pesticide detection In-line process monitoring |
| Sumario: | Pesticides play a crucial role in boosting the overall yield and productivity of agricultural produce by controlling pests, insects, and various plant diseases. However, excessive use of pesticides has led to contamination of food products and water bodies, as well as disruption of ecological and environmental systems. Global health authorities have set limits for pesticide residues in individual food items to ensure the availability of safe foods in the supply chain and to assist farmers in developing optimal agronomic practices for crop production. In Spain, specifically regarding olive cultivation, the Ministry of Agriculture, Fisheries, and Food establishes a safety period that farmers must observe from the application of the pesticide until the fruit is harvested. This period ensures that the batch of olives will comply with the maximum residue level allowed. This article proposes a methodology based on hyperspectral imaging to detect whether the olives have been sprayed with pesticide products and, if so, when the spraying occurred. The proposed methodology operates at the pixel level, where each pixel of the hyperspectral image is an instance. The pesticides evaluated were Diflufenican, Oxyfluorfen, Deltamethrin, - Cyhalothrin, and Tebuconazole. The results are promising and the success rates achieved over 80% accuracy for most pesticides in controlled laboratory conditions, with individual performance varying according to each pesticide’s chemical properties and stability on the olive surface. While the results are promising, the scalability of this approach for larger and more diverse batches of olives requires validation under field conditions, where variations in environmental factors, olive variety, and ripeness may impact the detection accuracy. Furthermore, the study highlights key wavelengths around 750 nm and 550 nm as effective discriminators, suggesting potential for cost-effective, simplified imaging systems. Although hyperspectral imaging shows potential as an accessible, in-line monitoring solution for cooperative use, further analysis of implementation costs is recommended to confirm its feasibility on an industrial scale. |
|---|