Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
Analysis of Broken Rice Kernels Using an Android Application
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.orgCitation: Applied Engineering in Agriculture. 38(2): 401-408. (doi: 10.13031/aea.14716) @2022
Authors: Karthik Salish, José Alfredo Gamboa, Kingsly Ambrose
Keywords: Android application, Computer vision, Image processing, Rice quality.
An android application was developed to analyze broken rice kernels. Maximum length and aspect ratio were good indicators to categorize broken rice kernels. The developed algorithm had a precision of 95.9% and an accuracy of 98.0%.
An android application was developed to analyze broken rice kernels.
Maximum length and aspect ratio were good indicators to categorize broken rice kernels.
The developed algorithm had a precision of 95.9% and an accuracy of 98.0%.
Abstract. The morphological characteristics of grain kernels play an important role in identifying the quality of rice. Manual sorting and inspection of rice kernels is a laborious process and susceptible to human errors. Mechanical separators such as indented cylindrical separators have also been used to separate broken kernels. In recent times, computer vision through image analysis has been applied to automate these processes, however, this necessitates image acquisition and processing devices. This article focuses on the development and use of an android application to determine the physical quality of rice kernels by quantifying broken grains using image processing and analysis techniques. The algorithm for the application includes several steps within image processing such as: image acquisition, preprocessing, segmentation, morphological transformation, and feature extraction. This quality inspection system was evaluated for medium-grain white rice. Experimental results showed a maximum average error of 2.8% in the prediction of broken kernels. This application can be used by primary producers and traders for analyzing the quality of rice.(Download PDF) (Export to EndNotes)