Moisture content is a crucial parameter in many industries:

  • For the food industry, this parameter strongly correlates the product quality. For instance, it is an indicator of freshness in the meat industry. It also needs to be monitored appropriately when one is producing dry fruits.
  • Monitoring the moisture level of fuel (e.g., wooden ships) to optimize the burning process and efficiency.
  • In the paper industry, water content needs to be monitored at many production stages to ensure quality.
  • In precision agriculture, moisture in leaves is a good indicator of plant health.

Monitoring the moisture content would thus increase production quality in many industries and optimize processes. New agile and accurate technologies are needed, and within this context, Specim cameras present many assets.

Water has strong absorption features within the NIR spectral range, and the use of Specim FX17 is therefore natural and evident for monitoring its content. NIR spectroscopy combined with chemometric algorithms reveals the moisture level in a quantitative manner. Since spectral imaging combines spectroscopy with imaging, the FX17 would also map the spatial distribution of moisture, which is crucial for some applications (e.g., precision agriculture and meat processing).

In this case study, we investigated minced meat samples’ moisture level. The moisture level correlates tightly with freshness, and its level needs to be precisely monitored, especially before packaging. For this study, Atria (Kauhajoki, Finland) provided ten minced meat samples. To know their moisture level accurately, Specim ordered measurements from a 3rd party laboratory, certified in moisture analysis (Seilab in Seinäjoki, Finland; method NMKL 14:2012; See Table 1 below).

Measured value by SeilabMeasured value by FX17
Sample 175.876.4
Sample 272.672.3
Sample 3 *69.269.5
Sample 464.764.3
Sample 5 (mostly fat)17.917.8
Sample 674.974.7
Sample 772.871.9
Sample 868.268.8
Sample 9 *60.362.2
Sample 1058.959.7
Table 1: moisture level (g/100g) on each sample included in this study. Samples 3 and 9 were used for validation purposes.

We measured the samples with Specim FX17 hyperspectral camera (Fig.1). It collects NIR spectra for each pixel of the acquired image (900 – 1700 nm). Those can be converted into moisture content employing a regression model. We used eight of these samples to build and calibrate the model. We used the two remaining samples for validation (indicated with * in Table 1). We used perClass Mira software to process the data.

Figure 1: FX17 on the 40×20 scanner (left) and example of a sample on the scanner sample tray (right).

The regression model results are presented in Table 1 and Fig.2. It clearly shows that the FX17 is a suitable tool to precisely measure minced meat’s moisture level.

Figure 2: regression plot of the quantitative model for moisture level prediction. Red dots relate to calibration samples, whereas green ones relate to validation samples.
Figure 2: regression plot of the quantitative model for moisture level prediction. Red dots relate to calibration samples, whereas green ones relate to validation samples.

In addition to measuring the moisture level in samples, hyperspectral imaging is suitable for measuring its distribution (Fig.3).

Figure 3: Example of Moisture distribution on a meat sample (here Sample 3).
Figure 3: Example of Moisture distribution on a meat sample (here Sample 3).

Conclusions:

The use of FX17 would provide integrators with crucial and accurate information for moisture level quantification. Besides, this fast and non-destructive method is also suitable to measure other properties, beneficial for process optimization or quality control. Similarly, the method’s flexibility allows a rapid adaptation to new regulations and challenges.