Contaminant-detection-in-packaged-cocoa-hazelnut-spread-jars-using-Microwaves-Sensing-and-

2400 instances of scattering parameters measured at 10.0 GHz to detect contaminants in packaged food jars. The dataset includes 1200 uncontaminated and 1200 contaminated samples across 11 classes of contaminants like metal, glass, and plastic. It was created by Luca Urbinati as part of a series of five datasets measured at different microwave frequencies.

Use Cases

• Training machine learning models for contaminant classification based on 30 scattering parameters.
• Developing anomaly detection systems for packaged food using microwave scattering data.
• Comparing the effectiveness of different microwave frequencies for detecting specific contaminant types and placements.
• Benchmarking sensor fusion or signal processing techniques for industrial food safety applications.

Strengths

• Contains 2400 total instances with a balanced 1200/1200 split between contaminated and uncontaminated samples.
• Includes 11 distinct, physically varied contaminant classes with specific instance counts (e.g., 200 for 'air_surface', 100 for 'glass_fragment_middle').
• Data is part of a controlled series acquired at five specific microwave frequencies, allowing for frequency-dependent analysis.

Limitations

• Column-level documentation is absent; the exact meaning and structure of the 30 scattering parameters must be inferred after download.
• Last update date is unknown; freshness unverified.
• The dataset's applicability is limited to the specific experimental setup (cocoa-hazelnut spread jar, specific contaminants, and 10.0 GHz frequency).

Provenance

Source

OpenML, author Luca Urbinati.

Collection Method

Microwave sensing measurements of scattering parameters from a network of antennas, a target jar (with/without contaminant), and air.