Implementation of Edge Detection Using the Sobel Operator on Papaya Leaf Images

Abstrak

Recent advances in digital image processing and computer vision have enhanced feature extraction techniques for plant identification based on leaf morphology. Edge detection is a fundamental operation that highlights intensity discontinuities corresponding to object boundaries. This study implements the Sobel operator to perform edge detection on tropical leaf images using an experimental–computational approach. The workflow involves grayscale conversion, horizontal and vertical Sobel filtering, and gradient magnitude computation implemented in Python using the OpenCV library. Experimental evaluation demonstrates that the Sobel operator effectively delineates primary leaf contours and preserves morphological consistency, despite reduced performance under non-uniform illumination and noisy conditions. These results confirm that the Sobel operator remains a reliable preprocessing technique for leaf-based feature extraction and classification, offering a computationally efficient baseline for future integration with machine learning-based plant recognition systems.