Made Windu Antara Kesiman, S.T., M.Sc., Ph.D.
Jam 13.00 WITA

The aim of the COFFECTION Project is to explore digital image processing methods and machine learning schemes to identify and to classify the defects types of coffee bean. Specifically, the aim of the project is to design and to develop an automatic grading system software by determining the defect value of coffee beans based on SNI 01-2907-2008, issued by National Standardization Agency (BSN). In this research, the device for capturing coffee bean images and the method for building ground truth images of defective coffee beans were built. To support our method, the construction of a dataset for classifying defective coffee beans was done. The capabilities of two different CNN models, MobileNet and InceptionResnet, to classify types of coffee bean defects have been explored and tested. COFFECTION software was built to identify the coffee bean defect type and to determine the defect value of coffee beans. A comprehensive software development process following the stages of the Software Development Lifecycle Model (SDLC Model) is carried out. Based on the results of testing and evaluation, the COFFECTION software has been able to fulfill all software requirements specifications both functionally and non-functionally. The interoperability design between the main COFFECTION engine and the web services for the COFFECTION interface also worked well.

Kami menghadirkan informasi seputar kegiatan akademik, program unggulan, inovasi, dan prestasi dari civitas akademika Undiksha. Jangan lupa untuk like, share, dan subscribe agar tidak ketinggalan konten menarik lainnya. Bersama Undiksha, kita wujudkan pendidikan unggul dan berdaya saing global!

#Undiksha
#GatraGanesha
#SalamHarmoni
#BeritaKampus
#undikshajaya

Apabila membutuhkan layanan informasi dapat menghubungi humas Universitas Pendidikan Ganesha melalui layanan https://go.undiksha.ac.id/kontak

Dapatkan informasi terkait Layanan Informasi, Sistem, Website, dan Update Terkini TI Undiksha melalui laman https://linktr.ee/upttik_undiksha