Gray-Level Co-Occurrence Matrix and Geometric Shape for Classification of Rubber Tree Maturity

Rubber trees can be said to be mature by having a trunk circumference of more than 45 cm at the height of 130 cm from the ground. It influences the use of digital imagery to determine the classification of “mature rubber trees” and “immature rubber trees”. The challenge in the image of rubber trees is that they have similar colour characteristics between tree trunks and the ground, and multi-object rubber trees in one picture. We propose a method using the gray-level co-occurrence matrix (GLCM) and geometric shape for the classification of Rubber Tree Maturity. GLCM is used to measure neighbouring pixels with grey intensity, distance, and angle in solving the first characteristic problem. Geometric shapes used to solve the second characteristic problem with the help of determining a region of interest (ROI). The research results show that the proposed method was successfully carried out with the strongest evidence on the support vector machine (SVM), namely 0.800 f1-score, 0.778 precision and 0.824 recall.