A comprehensive dataset of tomato leaf images for disease analysis in Bangladesh
Mangsura Kabir Oni1, Tabia Tanzin Prama1
1Jahangirnagar University, Dhaka 1342, Bangladesh.
Abstract:
Agriculture is the largest employment sector in Bangladesh, making up 13.4 percent of Bangladesh's GDP in 2024. For being most consumable crops, almost 9 out of 10 farmers grow tomatoes and earn their living. Tomato (Solanum lycopersicum) ranks fourth in respect of production and third in respect of area in Bangladesh. The tomato, a cornerstone of global agriculture, relies heavily on the health of its leaves for optimal growth and yield. These leaves are essential for photosynthesis, respiration, and transpiration, processes that directly influence the plant's overall vitality. Understanding the structure, function, and physiological characteristics of tomato leaves is crucial for developing effective agricultural strategies to maximize production and minimize the impact of environmental stressors and diseases. While tomato leaves exhibit a wide range of morphological variations across cultivars, they remain susceptible to a variety of threats. Pests, pathogens, nutrient deficiencies, temperature extremes, and environmental pollutants can all compromise leaf health. Biotic stresses, especially foliar diseases caused by bacteria, fungi, viruses, and other pathogens, are particularly devastating to tomato production. This research presents a dataset of leaves from tomato plants that are both insect-damaged and healthy. Our dataset contains 1,028 images of tomato leaves collected in Bangladesh, including 482 images of healthy leaves and 546 images of diseased leaves. The images were captured from two different tomato gardens in February, 2024 [1]. We captured the images the in diverse backgrounds, angles, and lighting conditions. Each image is precisely annotated to mark regions as either healthy or diseased, accounting for the complex background in each image. This dataset serves as a comprehensive resource for researchers and learners to analyze and improve the health management of tomato plants through the development of advanced computational models.
