Detection of Bacterial Speck Disease (Pseudomonas syringae pv. tomato) in Tomato Plants Grown in Greenhouses Using Image Processing

The quality and productivity of tomatoes depend on careful monitoring and timely interventions during cultivation. This study aims to determine the level of necrosis caused by bacterial speck disease (Pseudomonas syringae pv. tomato, Pst) on tomato leaves using image processing techniques. Stake tomato seedlings were used in greenhouse experiments. After spraying the pathogenic bacteria on the leaves, they were kept at 25–27 °C and 70–80% humidity until typical disease symptoms such as chlorosis and necrosis appeared. Disease diagnosis was initially conducted using OpenCV, a Python library capable of image processing and computer vision tasks. Based on the analysis of leaf images, disease severity and observation intervals were determined. Disease percentages were identified as follows: Scale 0 at 0.00%, Scale 1 at 8.64%, Scale 2 at 24.94%, Scale 3 at 27.78%, Scale 4 at 62.97%, and Scale 5 at 89.69%. It was observed that as disease detection rates increased, accuracy rates also rose, and standard deviation decreased. Although a slight increase was observed at Scale 3 due to environmental variation, the standard deviation decreased from 0.02695 at Scale 1 to 0.02131 at Scale 5. The algorithms used accurately detected bacterial specks as disease severity increased, reducing overall variability. The results of the greenhouse study suggest that early disease detection can mitigate product losses when applied in similar environments.