Can HyperSpectral Imaging prevent food wastage?

Sep 7, 20214 min read

By Ella Iwalewa U.

Flickr @ NOAA Great Lakes Environmental Research Laboratory

(A picture was taken by a hyperspectral camera to detect the formation of HypoAlgalBlooms in Lake Aeries. Photo Credit: Aerial Associates Photography, Inc. by Zachary Haslick)

First, what is hyperspectral imaging (HSI)? Hyperspectral imaging is the process of collecting and analyzing information from across the electromagnetic spectrum. In simple terms, it’s the process of observing a scene, image, or object closely to analyze the spectrum of each pixel of the image for detecting foreign objects, understanding chemical processes, and observation purposes.

In HSI, an imaging spectrometer is a device that’s responsible for capturing a spectrally resolved image (the detail of an image along the electromagnetic spectrum). The imaging spectrometer, through sensors, obtains information from a scene as a set of "images." These images are then grouped together to form what is called a hyperspectral data cube—a cube that comprises 2 spatial dimensions and 1 spectral dimension—to examine and process the information obtained from each image. Like other spectral scanners, HSI makes the analysis of a scene more in-depth because it makes use of the whole electromagnetic spectrum for this analysis. For example, while human beings see visible light in the three main spectra, red, green, and blue, HSI breaks down this spectrum into even more bands, meaning it differentiates the full spectrum in the pixel of an image. Also, every object leaves unique prints on the electromagnetic spectrum which are called "Spectral signatures." This makes it possible for the HSI to recognize the same object in the future.

The four types of HSI include a spatial scanner, spectral scanner, snapshot imaging, and spatio-spectral scanner.

Due to HSI’s reliability, it’s employed in several fields such as geology and biomedicine to detect the wear-out of tissues and possible growth of cancerous cells, and food/agriculture to prevent food waste & poor quality of food.

According to the United Nations Food and Agriculture Organization, thirty percent, or 1.3 billion tons, of food is wasted globally. Forty-five percent of vegetables are wasted and a whole twenty percent of meat is wasted as well. This is a staggering number considering over forty-five percent of the global population is food insecure. While there have been several humanitarian measures to prevent food wastage, food waste continues to store high.

Even the use of radiative scanners in food production companies doesn’t alleviate this issue, but rather makes it worse by allowing food to be overexposed to heat and susceptible to becoming carcinogenic. HSI can prevent food wastage without altering the chemical composition of food—also known as ripeness, firmness, or shelf-life of food products—as well as endangering the food. In order words, the data produced from HSI would be non-invasive to the food as well as timely. During the study, HSI sensors collect information on a particular crop, process the information (chemical composition of the food in this case) to understand the correlation between its unique appearance and food quality. Also, due to the large amount of waste that occurs during food processing as a result of insects, stones, etc. living inside an object, HSIs are able to detect elements as tiny as 0.05 millimeters (Trust me, those flies don’t stand a chance). This is made possible by the ability of the crops to have unique fingerprints for identification and HSIs hyperspectral cube analysis. In Australia, for example, HSIs are being used to study the various types of grapes and detect the time they will be diseased. In ImpactVision, a Machine Learning company that specializes in food technology, HSI is capable of unearthing the internal quality and chemical composition also known as ripeness, firmness, or shelf-life of food products.

However, the future of HSI in food production still remains vastly unknown due to several factors. Implementing HSIs is quite expensive, because of its size and the special features it has. Then, the analysis of food types would require a large amount of data about all food types and specifications of each. This data, which may be existing, is quite large to gather and process. Another issue is, the data produced by the hyperspectral cube may be difficult for the manufacturers to understand since it's not yet translated to physical properties of food like color, texture, etc.

While the reach of HSI has definitely been expanding in recent years, lots of research is needed to determine the best way to implement it in food production and make it accessible and easy to use for everyone. Remember though, a large percentage of food waste occurs in households, so do what you can to prevent it!

Questions

What is Hyperspectral imaging? Hyperspectral Imaging is the process of collecting and processing information across the electromagnetic spectrum for observation and recognition purposes.
What is an imaging spectrometer? Imaging spectrometers are devices used in Hyperspectral imaging to analyze a scene, object, or image. The data that is produced by the imaging spectrometer is cubic and leads to the formation of the hyperspectral cube.
What are the methods of hyperspectral imaging? Spectral imaging, spatial imaging, snapshot imaging, and spatial-spectral imaging.
How do I prevent food waste?