<![CDATA[Segmentation is an important step in building a fruit-quality classification system. Previous research has shown the success of the Otsu Thresholding method for fruit image segmentation, but its application to Carolina Reaper chili images in Indonesia has not been carried out specifically. This research proposes the integration of the Otsu Thresholding method with morphological operations to improve the accuracy of Carolina Reaper chili image segmentation based on the ripeness level. The process starts with RGB image acquisition using a controlled camera, followed by red channel extraction as the segmentation input. The Otsu method is used to separate the object and background based on pixel intensity, resulting in a binary image that is enhanced through morphological operations, including dilation, imfill, and bwareaopen. The results show high accuracy, with averages of 99.85% (mature), 99.38% (almost mature), and 99.67% (raw). The average computation time is less than one second which shows the potential for real-time applications. This research contributes to the efficiency of technology-based postharvest processing of Carolina chili peppers]]>
Copyrights © 2024
