2022
DOI: 10.1016/j.compag.2022.107007
|View full text |Cite
|
Sign up to set email alerts
|

Joint optimization of autoencoder and Self-Supervised Classifier: Anomaly detection of strawberries using hyperspectral imaging

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
9
0
2

Year Published

2022
2022
2024
2024

Publication Types

Select...
8

Relationship

0
8

Authors

Journals

citations
Cited by 25 publications
(11 citation statements)
references
References 33 publications
0
9
0
2
Order By: Relevance
“…Hyperspectral cameras could be used to provide more spectral bands, which might be beneficial for detecting discolorations. Liu et al Liu et al (2022) have demonstrated the effectiveness of combining hyperspectral imaging with advanced data analysis techniques, such as autoencoders and self-supervised classifiers, for anomaly detection in food items like strawberries. Similarly, hyperspectral imaging has proven useful for the early detection of mold in food products, as shown by Farrugia et al Farrugia et al (2021) .…”
Section: Resultsmentioning
confidence: 99%
“…Hyperspectral cameras could be used to provide more spectral bands, which might be beneficial for detecting discolorations. Liu et al Liu et al (2022) have demonstrated the effectiveness of combining hyperspectral imaging with advanced data analysis techniques, such as autoencoders and self-supervised classifiers, for anomaly detection in food items like strawberries. Similarly, hyperspectral imaging has proven useful for the early detection of mold in food products, as shown by Farrugia et al Farrugia et al (2021) .…”
Section: Resultsmentioning
confidence: 99%
“…El desarrollo de métodos de detección de anomalías no supervisados para datos hiperespectrales es de gran importancia para sus aplicaciones en control de calidad y seguridad (Tabla 1). Como método de detección de anomalías de uso frecuente, el codificador automático puede sufrir la ineficacia de extraer representaciones esenciales para distinguir muestras normales y anómalas, ya que solo está entrenado para minimizar el error de recons-trucción (Liu et al, 2022). Recientemente, la detección automática de arándanos podridos sigue siendo un desafío en la industria alimentaria.…”
Section: Evaluación De Anomalías Enfermedades Y Daños Mecánicos En Di...unclassified
“…La descomposición temprana de los arándanos ocurre en la cáscara de la superficie, que puede adoptar la viabilidad del modo de imagen hiperespectral para detectar la región de los arándanos en descomposición (Qiao, Zhang, & Pei, 2020). En un estudio realizado por Gao, Zhao, Hoheisel, Khot, & Zhang (2021) El SSC-AE se degradó correctamente y superó todos los métodos de comparación en todos los niveles de impurezas (Liu, et al, 2022).…”
Section: Evaluación De Anomalías Enfermedades Y Daños Mecánicos En Di...unclassified
“…Mae is the mean of the absolute error and it denotes the error of the predicted values. Mdae is calculated as the loss relative to the median value for all the absolute differences between the observed and predicted values, and provides a measure of the robustness of the variances (An et al, 2022;Liu et al, 2022;Pham & Liou, 2022). Equations 12-17: In Equations 12-17, n denotes the number of samples, y act is the observed value, y pred is the predicted value, and y mean is the mean of the measured values.…”
Section: Fitmentioning
confidence: 99%
“…Hyperspectral imaging combines the spectrum and image of the target object at the same time to accurately capture the spectral data and image information of each pixel in the image (Liu et al, 2022). In recent years, hyperspectral imaging and visualization techniques have been applied in agriculture for drought monitoring and the control of diseases and insect pests (Sun et al, 2019).…”
Section: Introductionmentioning
confidence: 99%