How To Capture Spectre Color Data For Comparison
Objective: Collect and store red apple bin image data, to allow the Spectre team to validate and improve the color estimation process.
Note: The same data can be used to run size comparisons.
Options for red apple bin sets to prepare for scanning:
- Pull aside a set of 10 - 60 red apple bins that can be graded together to produce a single report
- If Option 1 is not possible, prepare an entire lot/grader run for scanning
- If Option 2 is not possible, prepare a subset of 10 - 60 apple bins from a lot/grader run for scanning
Steps to capture the images with the correct data allocation:
- Open Spectre Group
- Click Create a group
- Select the correct apple bin dimensions, pre-loaded into Spectre
- Select the correct variety
- Select the correct lot
- Select the correct block (if applicable)
- Type ‘Color Test’ in the TruckID field
- Capture the apple bin images, one at a time. Note that there is a maximum of 30 images per group. If the set of bins pulled aside exceeds 30, another group(s) will be required.
- Upload the group
- Review the group in ‘Today’s Results’ to ensure that the images were successfully processed (optional for well practiced users)
Providing the grader data for the set of images:
- Commit the set of apple bins captured into the grading machine
- Produce a grader report, that clearly outlines in terms of pieces and/or weight, the color and size grades seen in the set of apple bins
- Ensure that we’ve received a breakdown of the color grade thresholds for each given variety. For example for Variety 1:
- Grade 1 = 75% - 100% red
- Grade 2 = 45% - 75% red
- Grade 3 = 20% - 45% red
- Grade 4 = 0% - 20% red
- Inform Elizabeth & Luke once this information has been captured
Thank you for joining us on the journey to building the world's most user friendly fruit quality estimation tool!