Application Of The Template Matching Method In Orchid Image Classification In Bengkulu

Abstract

Bengkulu is one of the provinces in Indonesia, precisely located in the western part of Indonesia, namely on the island of Sumatra. Bengkulu has a wide variety of beautiful flowers, including orchids which are unique to Bengkulu. Bengkulu orchids are included as rare and endangered flowers, and the lack of public knowledge of the shape and characteristics of these flowers makes the author decide to create a system that can identify the type of orchid. In classifying the image texture, it can be done using the template matching method. The template matching method is a simple method used to identify patterns in an image. The template matching method is often used to identify character images of letters, numbers, fingerprints and other matching applications. In this study, we propose the use of a template matching algorithm to classify orchid images based on species and family. The test results use 30 test image data. The minimum correlation value with the correct label on the orchid image is 0.7967 and the classification image is 0.7899. From the experimental results it can be concluded that the template matching algorithm is quite good in orchid image classification as evidenced by a test accuracy of 76 percent.