Research Fellow in Image Processing, Media and Game Innovation Centre of Excellence (MaGICX), University Technology Malaysia, Skudai, Malaysia

Biograph
Mohd Shafry Mohd Rahim, PhD in Spatial Modelling from University of Agriculture Malaysia, Malaysia in the field of Geographical Information Systems and Computer Vision for agriculture. He focuses on Image Processing in various area of application. His current research focus is on plant identification based on botanical knowledge and rubber tree identification. Besides, he also works on Computer Vision and Internet of Things (IoT) for Agriculture.

Speech Title: An expert botanical feature extraction technique for identifying plant species

Abstract: Automated Plant Species Identification System is important application for agriculture industry and other fields. It’s needed to better understand of their use and preserve the biodiversity. In this presentation, we present a new method to recognize the leaf and identify plant species using some unpopular parts of the leaf including; lobes, apex and base detection. Most of the researches in this area focus on popular features such as the shape, color, vein and texture which consume large amounts of computational processing and which are not efficient, especially in Acer database with a high complexity structure of the leaves. This method is focused on some unpopular parts of the leaf which increases accuracy. Detecting the local maxima and local minima is done based on Centroid Contour Distance for Every Boundary Point, using north and south region to recognize the apex and base. Digital morphology is used to measure the leaf shape and the leaf margin. Centroid Contour Gradient is presented to extract the curvature of leaf apex and base. We analyze 32 leaf images of tropical plants and evaluated with two different datasets, Flavia, and Acer. The best accuracy obtained is 94.76% and 82.6% respectively. Experimental results show the effectiveness of the proposed technique without considering the popular features with high computational cost.

Keywords: Botanical Feature Extraction, Leaf recognition, Plant Species Detection, Leaf Apex, Leaf Base