Science and Research Journal
|ISSN 2619-9955 | eISSN 2619-9963 | Frequency:4/year | Peer Reviewed: Yes | UIKTEN Publisher|
Combination of Chain Code and Leaf Morphology Approach for Plant Identification
Oktaviani Oktaviani, Sarifuddin Madenda, Rodiah Rodiah, Diana Tri Susetianingtias, Fitrianingsih, Dea Adlina, Rini Arianty
© 2019 Rodiah Rodiah, published by UIKTEN. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. (CC BY-NC-ND 4.0)
Citation Information: SAR Journal. Volume 2, Issue 3, Pages 89-95, ISSN 2619-9955, DOI: 10.18421/SAR23-01, September 2019.
Received: 25 July 2019.
Revised: 01 September 2019.
Accepted: 07 September 2019.
Published: 28 September 2019.
This research will identify plants based on 16 basic leaf shapes. It uses 3 leaf datasets, namely Flavia, Folio and Swedish dataset. The researcher used 12 features that consist of 5 basic geometry features and 7 digital morphological features. Geometry features include physiology length, physiology width, diameter, area, and leaf perimeter while digital morphological features involve smooth factors, aspect ratio, circularity, rectangularity, narrow factor, the ratio between perimeter and diameter combined with chain code in determining the ratio between perimeters and lastly, the length and width of physiology. Accuracy result showed that 98.48% of the plants were correctly identified.
Keywords – Chain Code Extraction, Feature, Leaf.