Analysis of leaf morphological features for identification of medicine plants using support vector machine

Indonesia is a country of biodiversity with a wealth of medicinal plants that have the potential to be developed. The number of medicinal plants and the lack of public knowledge about the types of medicinal plants makes it difficult for people to distinguish them, so many people prefer to use chemical drugs. Based on the many types of medicinal plants, the lack of public knowledge about medicinal plants, and changes in people’s lifestyles in consuming medicinal plants, a system is needed to identify the types of medicinal plants. Leaves are used to identify types of medicinal plants, because each type of medicinal plant has a different leaf shape. This study uses leaf morphology patterns as objects to identify types of medicinal plants. Analysis of leaf morphological features will produce values of morphological characteristics, which can then be processed using a computer to identify medicinal plant species by applying the Support Vector Machine algorithm. The results obtained from the SVM method are able to identify medicinal plants with an accuracy of 85.5%. The SVM method is very good at identifying binahong leaves and som Jawa. Errors occur because the variation in medicinal plant data is very small.