Garden Flowers and Butterfly Interaction Image Recognition Dataset

Currently, the demand for automated recognition systems in agriculture and forestry is increasing, especially in horticultural management and natural landscape monitoring. However, existing image recognition technologies still face challenges in accurately recognizing specific flower species and butterfly behaviors. This dataset provides high-quality images of flower and butterfly interactions, offering necessary data support to enhance the accuracy and efficiency of automated recognition systems. During data collection, we used high-resolution cameras to capture images of flowers and butterflies in various garden environments on sunny days. Quality control includes multiple rounds of professional annotation and expert team reviews to ensure data consistency and accuracy. The annotation team consists of 10 experts with backgrounds in botany and entomology. Data preprocessing uses image enhancement and noise filtering techniques, and the final data is stored as structured files in JPG format for easy retrieval and analysis. The data features high annotation precision and completeness, with an innovative multi-level annotation approach to improve data consistency and coverage. Novel image enhancement techniques further ensure the dataset's performance under different lighting conditions. Compared to similar datasets on the market, this dataset boasts 15% higher annotation accuracy, providing strong support for improving the accuracy of automated recognition algorithms, especially in complex horticultural environments. The data can be widely applied in plant protection and species monitoring in agricultural automation systems and can be extended to other similar biometric applications, offering high versatility.