A Field-Collected Multiclass Marigold Flower Image Dataset from Bangladesh for Cultivar Classification and Computer Vision Applications

**Description** The Marigold-5 dataset is a field-collected image dataset containing photographs of five marigold cultivars captured under natural environmental conditions in Bangladesh. The images were collected from three geographically distinct locations: * Bogura (B-Block Area) * Mymensingh * Daffodil International University Campus, Dhaka The dataset was developed to support research in computer vision, machine learning, image classification, agricultural informatics, flower cultivar recognition, and plant phenotyping. The dataset includes images of the following marigold cultivars: 1. African Marigold (Orange) 2. African Marigold (Yellow) 3. French Marigold (Red Cultivar) 4. French Marigold (Red-Yellow Cultivar) 5. Hybrid Marigold * American Hybrid * French Hybrid All images were captured directly from marigold plants under real-world field conditions using natural lighting and diverse backgrounds. The dataset contains raw, unprocessed images without augmentation, synthetic modifications, or artificial enhancement. Variations in viewpoint, illumination, flower maturity, and environmental conditions were intentionally preserved to reflect real-world cultivation scenarios. This dataset can be used for flower cultivar classification, image recognition, machine learning, deep learning, transfer learning, object detection, and artificial intelligence-based agricultural applications.