What does this research mean for the field?

A non-invasive computer-vision method using deep learning can reliably detect water adulteration in orange juice with high sensitivity and specificity, serving as an effective rapid pre-screening tool. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To develop a rapid, non-invasive computer vision and deep learning framework to detect and classify water adulteration levels between 1% and 15% across commercial orange juice products.

September 3, 2026Journal of Food Composition and AnalysisOpen Access

A Non-Invasive Approach for Detecting Water Adulteration in Orange Juice Using Computer Vision and Deep Learning

View Full Paper

Ask AI

Bookmark

Authors

APAna M. Pérez-Calabuig SPSandra Pradana‐López JCJohn C. Cancilla

Discussion

Member takes

Overview

Experimental study reveals deep learning accurately detects water adulteration in commercial orange juice products, indicating high feasibility for rapid beverage authentication.

Key Points

To develop a rapid, non-invasive computer vision and deep learning framework to detect and classify water adulteration levels between 1% and 15% across commercial orange juice products.
Trained ResNet50 convolutional neural networks using high-resolution images of fresh-squeezed juice, juice from concentrate, and orange nectar captured at two shutter speeds (1/30 s and 1/250 s).
Evaluated sample-level transferability through an independent blind validation on N=240 newly prepared samples obtained from subsequent retail purchases.
The 1/250 s exposure model achieved 88.3% accuracy in hold-out testing (compared to 83.1% for the 1/30 s model) and attained 86.7% 24-class accuracy during independent blind validation.
In a binary screening configuration (pure versus adulterated), the model achieved 100.0% sensitivity, 93.3% specificity, 99.2% accuracy, and 96.7% balanced accuracy.

Cite This Study

Pérez-Calabuig et al. (2026) studied this question.

synapsesocial.com/papers/6a9935f3636c6408cfa7e9dc https://doi.org/10.1016/j.jfca.2026.109484

View Full Paper

Ask AI

Bookmark

September 3, 2026Journal of Food Composition and AnalysisOpen Access

A Non-Invasive Approach for Detecting Water Adulteration in Orange Juice Using Computer Vision and Deep Learning

View Full Paper

Ask AI

Bookmark

Authors

APAna M. Pérez-Calabuig SPSandra Pradana‐López JCJohn C. Cancilla

Discussion

Member takes

Overview

Experimental study reveals deep learning accurately detects water adulteration in commercial orange juice products, indicating high feasibility for rapid beverage authentication.

Key Points

To develop a rapid, non-invasive computer vision and deep learning framework to detect and classify water adulteration levels between 1% and 15% across commercial orange juice products.
Trained ResNet50 convolutional neural networks using high-resolution images of fresh-squeezed juice, juice from concentrate, and orange nectar captured at two shutter speeds (1/30 s and 1/250 s).
Evaluated sample-level transferability through an independent blind validation on N=240 newly prepared samples obtained from subsequent retail purchases.
The 1/250 s exposure model achieved 88.3% accuracy in hold-out testing (compared to 83.1% for the 1/30 s model) and attained 86.7% 24-class accuracy during independent blind validation.
In a binary screening configuration (pure versus adulterated), the model achieved 100.0% sensitivity, 93.3% specificity, 99.2% accuracy, and 96.7% balanced accuracy.

Cite This Study

Pérez-Calabuig et al. (2026) studied this question.

synapsesocial.com/papers/6a9935f3636c6408cfa7e9dc https://doi.org/10.1016/j.jfca.2026.109484

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark