Garden Flower Calendula (Marigold) Recognition Image Dataset

With the development of urban horticulture and ecological gardens, precise identification of flower species has become increasingly important in horticultural management and botanical research. Currently, manual recognition and traditional image processing methods often face issues such as insufficient accuracy and difficulty in distinguishing species, especially in complex natural environments. Existing solutions are limited in terms of species diversity and environmental adaptability. This dataset aims to improve the accuracy of Calendula recognition in natural and varied conditions using deep learning technology, aiding horticultural management and research. Data collection is primarily conducted using high-definition digital cameras under various natural light conditions to ensure diversity and authenticity. Quality control includes multiple rounds of annotation and expert consistency review, with the annotation team consisting of botanists and data analysts, totaling 20 members. The data is preprocessed with noise reduction and color enhancement, stored in a layered directory structure in JPG format for easy indexing. Core advantages include a labeling accuracy of up to 95% and good annotation consistency, innovatively employing an ambient spectral segmentation enhancement algorithm to address accuracy issues under multiple lighting conditions. The dataset can effectively enhance the recognition performance of flower management systems. Compared to other single-species datasets, its diversity and environmental coverage are absolute advantages, especially suitable for expansion to other flower species recognition.