A New Machine Learning Method for Identifying Fake Honey

A New Machine Learning Method for Identifying Fake Honey

A researcher’s team at Imperial College London and UCL has developed a method for authenticating machine learning. This technique that can detect mislabelled honey was published in the paper published on the arXiv.

Honey is one of the fabricated food products that can be mislabelled and can lead to selling wrong products or diluted product.

Gerard Glowacki, one of the researchers said that honey plant-derived product, which contains different pollens. However, the fake honey costs less but it has a low amount of pollens. In addition, numerous people have the practice to sell fake honey often compared to practices in honey farming, which can mistreatment with bee colonies.

The Melissopalynology has been around for a long time, but it works slowly. The new technique has the intention to fasten up the process and avoid suffering from human things forgetfulness.

Other most common authentication methods include measurable polymerase chain reaction, liquid chromatography-mass spectrometry, nuclear magnetic resonance spectroscopy, and microscopy. Researchers have developed other parameters for the identification of specific types of honey. The current technique is able to authenticate some type of honey; however, it is not applicable to other types of honey.

Mostly honey authentication procedures conduct tests in labs and specialists and require expensive some equipment. The researchers developed a new method for authenticating machine learning and augmented microscopy.

After labeling and images, the researchers used for training and machine learning model on honey. The model consists of the segmentation network and segmenting pollen and training for sorting different types of honey.

Leave a Reply

Your email address will not be published. Required fields are marked *