VegNet: Vegetable Image Dataset

This dataset comprises 6,244 high-resolution JPG images of 40 distinct vegetable varieties, collected from street stalls and local markets in Mirpur, Dhaka City, Bangladesh. Captured under natural lighting conditions using a Smartphone HD camera, the images accurately represent each vegetable’s unique visual traits—including color, texture, and shape—for reliable categorization and analysis. The dataset spans diverse varieties such as Arum Lobe, Ash Gourd, Bitter Melon, Eggplant, Tomato, and Yardlong Bean, among others, ensuring comprehensive coverage. To facilitate computational analysis, the dataset includes both original and resized versions of the images, optimized for machine learning and computer vision applications. Collected between Jan 1 and March 15, 2025, this resource supports researchers, agronomists, and data scientists in advancing vegetable classification, agricultural technology, and food security initiatives. By providing a balanced and detailed visual repository, the dataset bridges the gap between real-world variability and AI-driven agricultural solutions.