It has been a while since I used one of these, so I don't recall the different head-piece options, but is it possible to measure more than one fruit?

I gather that if you only have a single cherry for example in the field of view, the average light colour response will be skewed towards the white background much more say than an apricot or apple.

If you place many fruits in the field of vision, even 2 or 3 deep, you will probably get a better "average", but with the side effect of a lot less precision about that average. This would mean taking many samples per lot (eg a kilogram of cherries sampled 10-20 times with suitable re-randomising of fruit each time). If you plot the coefficient of variation vs number of samples, you can see how many subsamples you need before the precision is good enough for your study purposes. Larger fruits will have greater depth issues, but fewer subsampling errors.

I recall somewhere reading about a "neutral grey" plate for small objects, but I don't know if that was for this purpose or not.

Have you read through the manual for any helpful info? https://sensing.konicaminolta.us/wp-content/uploads/cr-400_410_instructions_eng-t31g0h9a29.pdf

Check for details https://www.konicaminolta.com/instruments/download/instruction_manual/color/pdf/cr-400-410_instruction_eng.pdf

Also check https://www.researchgate.net/publication/225037588_Colour_Measurement_and_Analysis_in_Fresh_and_Processed_Foods_A_Review

You need to do color calibration using Xrite Color Checker Passport and Adobe Lightroom to produce a calibrated tiff file. Then the image background is removed using quick selection tool and color decontamination tool from an image processing software (Adobe Photoshop CC 2015) and the resulting image is saved into a tiff file.

Thank you Matthew Wheal for your helpful answer. I would like to have more information about the "neutral grey" plate, have you any documents about it?

Duried

No I don't, sorry, it was over 20 years ago. I just recall that it was some sort of background colour at 18% grey or a similar concept, which had minimal red green or blue. It may have been related to measuring light intensity rather than colour profiles. It still doesn't stop the average colour being influenced by the background of the plate. I liked Kiprotich Kiptum’s idea of editing the image with software, but that may not give an accurate reading direct from the minolta unit.