<![CDATA[Berries are known as "super fruits" because they are rich in nutrients and antioxidants. Despite their abundant health benefits and significant economic potential in Indonesia, classifying berries like blackberries, goji berries, and mulberries is often challenging due to their visual similarities. This challenge hinders the sorting process in the agricultural and trade industries. This research proposes an automated classification model based on image processing to identify these three types of berries this study aims to develop and test effectiveness of an automatic classification model using image processing to accurately differentiate between blackberry, goji berry, and mulberry, in order to address the difficulties of manual sorting in the industry. The study implements K-Means Clustering as the primary technique for image segmentation, aiming to accurately separate the fruit from its background. The workflow begins with collecting 30 images of each berry type, followed by a color space transformation from RGB to Lab to separate color and brightness components. After segmentation, shape and texture feature extraction is performed to obtain the unique characteristics of each fruit. The analysis results show that feature extraction successfully captured significant differences between the three fruits. Blackberries tend to be rounder (metric: 0.56934; eccentricity: 0.56594), whereas goji berries (metric: 0.15132; eccentricity: 0.92832) and mulberries (metric: 0.097072; eccentricity: 0.87125) are oblong. Texture analysis also shows that mulberries have the smoothest surface. These quantitative differences are key to distinguishing the three fruits. Overall, this method provides an effective and accurate identification solution that can be implemented in automated fruit sorting systems to improve the production quality and economic value of berries in Indonesia.]]>
Copyrights © 2025
