Written by Ines Hanrahan and Stefan Röder, Washington Tree Fruit Research Commission
The F-750 is a hand held device, which uses near infrared spectroscopy to determine different fruit quality parameters. This article provides an overview on the necessary steps for use of this device.
In order to use the F-750, you must develop a model specific to your site. The procedure includes five steps. The first three steps are shown in the figure below.
Step 1: Sample Preparation
- Select apples with a wide maturity range (for example 120 pieces).
- Label the samples with continuous numbers.
- Mark the measuring position on the sample with a circle and label the circle with a number (1-3). Note: we used three measuring positions per fruit (1 for a DA-meter and firmness model, 2 for a dry matter model and 3 for a SSC & TA model).
Step 2: Creating a Training Set
- During this step, each circle (measurement position) will be scanned with the F-750.
- First measure all 120 positions with the first set of circles facing the lens, then measure the second set of circles and then measure the third set of circles.
Note: the training set created in this example shown would be usable as a postharvest set only, since all measurements were performed at a single temperature
Step 3: Measuring Reference Values
- After creating a training set, measure your desired parameter (firmness, IAD, dry matter content, Soluble Solids Content and/or Titratable Acidity) using the standard methods.
- Make sure to always choose the same labeled circle for the specific measurement.
E.I. #1 for IAD and then for firmness, # 2 for dry matter and # 3 for the juice (SSC, TA).
Dry matter content
In order to measure the dry matter content, remove the peel from the area of interest. Then cut a slice of the apple and punch out a sample using the tools provided. Measure the fresh weight immediately and put the sample in the oven (at 70° C or 158° F) for 48 hours until a constant dry weight is reached. At the end, measure the dry weight and calculate the dry matter content.
Note: we have observed differences in values achieved with this method vs. a standard method for dry matter content used by WTFRC.
To measure the firmness, simply remove the peel from the measurement position and perform the measurement using a penetrometer.
Soluble Solids Content (SSC) & Titratable Acidity (TA)
Separate the area of interest by slicing the apple. Place the sample in a juicer and measure the SSC and TA content using a refractometer and a titrator.
Step 4: Building the Model
- Building the model requires some IT and statistical knowledge (Felix Inc.).
- After downloading and installing the software provided by Felix Inc., load the training set on your computer and open it with the model builder software.
- Transfer your reference values into the software.
- The software will look for a correlation in your data set, it also provides tools to analyze the performance of your model.
- The F-750 handbook includes a detailed description on how to build and analyze the models (The link is provided at the end of the document).
Step 5: Validating the Model
- To validate your developed models repeat step 1 to 3 with new samples and compare the results from the F-750 with your reference values.
- Instead of step 2 just load the new model onto the device.
Note: Before purchasing the device, the following considerations should be made: Currently it is not clear if there is a need for variety specific or pre- and postharvest models; scientific literature does not always show strong correlations to harvest maturity or storability between all variables measured and/or for all commercial varieties grown within different regions.The current software allows integration of up to three parameters in one model (although the instrument can read five parameters total). In order to use the F-750 in the field, the models have to be adjusted for different temperatures, therefore, the creation of a training set for field use will take considerably more time and samples.
Link to the manufacturer
Link to a detailed manual
Ines Hanrahan and Stefan Röder
Washington Tree Fruit Research Commission