<![CDATA[Traditional cakes or generally referred to as market snacks are one of the many traditional specialties originating from Indonesia and are usually traded in traditional markets. The cakes that are traded are of various types and have a distinctive taste and are not inferior to modern food. The traditional cake is usually packaged in an attractive and unique form, the wrapper can come from leaves that have fallen or are still alive to be wrapped in plastic in order to attract consumers or buyers. Even though in today's era, there are not a few instant and practical food products, even some imported products from abroad whose packaging is more attractive. And this traditional cake is one of many cultural heritages that should be more commensurate with other Indonesian cultural assets. Therefore, as the color of Indonesia, we should maintain and preserve and further introduce to all levels of society that traditional cakes are no less delicious than modern food, especially children today who are foreign to these traditional cakes. Therefore, a system is needed to identify traditional foods, especially traditional cakes, by utilizing the sophistication of technology that exists in the current digital era. This study proposes the application of image segmentation on traditional cakes using the DBSCAN algorithm to obtain cake image segmentation results with an average Intersection over Union (IoU) accuracy of 91.3% and a maximum value of 99.8%. This shows that the proposed method is able to provide the best results.]]>
Copyrights © 2021
