ANALYZING THE COMPARATIVE METHODS OF PREWITT, ROBINSON, KRISCH AND ROBERTS IN DETECTING THE EDGES OF RICE LEAVES

Abstract

This research explores the vital role of rice in Indonesia as a staple food and primary source of income for farmers. Efforts are being made to increase rice production to meet the growing demand. The study focuses on object edge detection in image analysis, evaluating methods like Prewitt, Robinson, Krisch, and Roberts. Digital imaging plays a crucial part in visually presenting information, and image processing improves image quality for human and machine recognition. Detecting object edges, particularly in rice leaf images, is essential for computer inspection. The experiment on fifteen rice leaf images shows that the Krisch method performs better in edge detection, with a 52% average accuracy and smoothness. Other methods, such as Prewitt (6%), Robinson (11%), and Roberts (14%), have lower accuracy rates. These findings provide a foundation for enhancing edge detection in rice leaf image analysis. The study also emphasizes the need for refining classification models. Overall, this research provides insights into the effectiveness of edge detection methods in rice leaf image analysis.