Submitted to: National Meeting of Institute of Food Technologists/Food Expo
Publication Type: Abstract Only
Publication Acceptance Date: 6/24/1998
Publication Date: N/A
Citation: N/A Interpretive Summary:
Technical Abstract: The peanut pod hull-scrape method is widely used to determine peanut pod maturity distribution and recommend harvest date for certain peanut cultivars and growing regions. A relatively large representative pod population sample is manually color sorted into pre-determined color classes. It is time consuming and susceptible to variability depending on subjective factors including individual preferences. Objective, automated color classification would increase the consistency of test results while also increasing the efficiency and documentation capability. The research objective was to design and test a computer-based image acquisition and analysis method and procedure for color classification of peanut pods based on the established hull-scrape methodology. A CCD color video camera with RGB output coupled to a personal computer equipped with an RGB video input/output card were used for image acquisition. Commercial color classification software was used to build color reference classes based on manually selected color classes. Color distributions were used to sort the test pods into sets of 50 by class comparison for a variety of pod arrangements and distributions. Computer generated classification was generally within 1 color class relative to manually sorted samples when whole pod surface was sampled as a color distribution. Classification of a 50-pod sample was performed in under 1 sec. The objective computerized color classification method could significantly improve the accuracy, repeatability and efficiency of hull-scrape methodology for peanut pod maturity determination. In combination with Internet communication capabilities, remote sensing and analysis systems could be implemented instead of the current practice based on manual sorting.