Berrylyzer - Berry Image Analysis System
The Food Science department at Oregon State University processes thousands pounds and hundred of genotypes of small fruits including blueberries, strawberries, red raspberries, black raspberries, and blackberries. The human-power it takes to process these is rigorous and an assistive technology is proposed. When people evaluate tons of berries they tend over time to judge “subjective” traits when the idea is to be objective. A computer-based vision system is proposed that is objective to the strict, rigorous color saturation, size, defect, and shape measurements that are needed to sort goods. Selecting and assessing berries based on the image processing system would add a helpful layer to the processing folks in food science! A software package that can assist in the speedy evaluation of visual characteristics of new berry breeding selections for small fruits research is needed.
Objectives
Objective: portable, dynamic, robust, image classification software package that takes in multiple image types and stores assessment in cloud database
Motivations
Hundreds of new selections of small fruits are evaluated for various quality characteristics each season by OSU and USDA researchers. Having a tool to objectively evaluate visual aspects of the fruit would assist in data collection and help to catalogue new breeding selections each season. Stretch goals can be berry identification and classification.
Qualifications
Minimum Qualifications:
Image processing
Database storage/manipulation
Like food!
Preferred Qualifications:None Listed
Details
Project Partner:
Zachary Wiegand
NDA/IPA:No Agreement Required
Number Groups:1
Project Status:Accepting Applicants
Website:https://foodsci.oregonstate.edu/foodsci/fst-facilities-pilot-processing-plant
Keywords:
Software DevelopmentImage ProcessingBerry BreedingBerry AnalysisVisual InspectionFood science