Authors :
Jocelyn B. Barbosa, Vergel I. Jabunan, Tatiana Kay A. Lacson, Ma. Lesley W. Mabaylan, Gera Mae M. Napone
Volume/Issue :
Volume 2 - 2017, Issue 10 - October
Google Scholar :
https://goo.gl/DF9R4u
Scribd :
https://goo.gl/1P3tSB
Thomson Reuters ResearcherID :
https://goo.gl/3bkzwv
Abstract :
Imaging technology has taken off at its significant level in the last decades. It has been used in different areas of research such are those that tackle plant recognition. In fact, there has been considerable body of work that performs analysis on leaf images, but most of them focus on plant or leaf identification. In this study, we present HerbApp, a mobile-based application that serves as a convenient tool in discriminating herbal from non-herbal plants to develop awareness among people on the significance of the plants whether or not it has been known publicly. Different characteristics and features of plants are used to perform pattern recognition and data analysis. From the captured leaf image, we perform segmentation process based on Localized Active Contour (LAC) model and extract features, which are used to build a classifier for leaf classification using Regularized Logistic Regression (RLR). Experiments show that our approach provides efficient results.
Keywords :
Leaf Recognition; Herbal Leaf Recognition; Herbal And Non-Herbal Discrimination; Medicinal Leaf Recognition; Image Processing; Regularized Logistic Regression; Data Mining; Local Active Contour Model; LAC-Based Segmentation.
Imaging technology has taken off at its significant level in the last decades. It has been used in different areas of research such are those that tackle plant recognition. In fact, there has been considerable body of work that performs analysis on leaf images, but most of them focus on plant or leaf identification. In this study, we present HerbApp, a mobile-based application that serves as a convenient tool in discriminating herbal from non-herbal plants to develop awareness among people on the significance of the plants whether or not it has been known publicly. Different characteristics and features of plants are used to perform pattern recognition and data analysis. From the captured leaf image, we perform segmentation process based on Localized Active Contour (LAC) model and extract features, which are used to build a classifier for leaf classification using Regularized Logistic Regression (RLR). Experiments show that our approach provides efficient results.
Keywords :
Leaf Recognition; Herbal Leaf Recognition; Herbal And Non-Herbal Discrimination; Medicinal Leaf Recognition; Image Processing; Regularized Logistic Regression; Data Mining; Local Active Contour Model; LAC-Based Segmentation.
