Optimisation of Image Morphology Operations with Enhancement and Convolution in Tomato Leaf Disease Symptom Recognition

Tomato (Solanum lycopersicum) is an important horticultural crop that is highly susceptible to various leaf diseases such as leaf spot, bacterial wilt, and fruit rot, which significantly reduce yield and quality. This study applies digital image processing techniques including pre-processing, morphology, enhancement, and convolution to improve the recognition of disease symptoms on tomato leaves. Pre-processing using grayscale conversion and median blur effectively reduces noise and sharpens essential details, while morphological operations (erosion and dilation) highlight structural features of infected areas. Enhancement techniques increase image contrast, making the distinction between healthy and diseased tissue more visible. Convolution methods with kernels such as Sobel and Gabor further emphasize edges and texture patterns of leaf lesions. Experimental results show that these methods improve pixel intensity distribution and enhance the visibility of disease symptoms, thereby increasing diagnostic accuracy. The integration of these techniques demonstrates the potential for early detection and classification of tomato leaf diseases, enabling more effective disease management and prevention of crop losses.