Lavender Recognition Image Dataset for Garden Flowers

Currently, the demand for plant recognition in agriculture, forestry, and fishery is increasing, but traditional recognition relies on manpower and has limited accuracy. Automatic recognition systems for specific plants like lavender are still immature, and existing solutions face challenges in handling large-scale diverse data. This dataset aims to address the accuracy and efficiency issues of lavender recognition, reduce labor costs, and enhance automation levels. Data is collected using high-resolution cameras to ensure diversity; it is gathered under different weather and lighting conditions to enhance generalization capabilities. Quality control involves multiple rounds of annotation and expert review to ensure consistency and high annotation accuracy. The annotation team consists of botany experts and data analysts, with a team size of over 10 people. Data preprocessing employs methods such as image enhancement, denoising, and normalization. The final data is stored in JPG format and organized by time and location for easy retrieval and analysis. This dataset boasts high annotation accuracy and consistency, with an annotation error rate of less than 1%. Utilizing new image enhancement and quality evaluation methods, it improves model generalization capabilities and recognition accuracy. Compared with other flower datasets, its uniqueness lies in the large and diverse amount of data specifically for the lavender variety, providing unique data support for lavender recognition and classification. Experiments have shown that using this dataset for training improves recognition accuracy by 20%, significantly outperforming other general flower datasets. Through adaptive enhancement technology, this set can be extended to recognition tasks of other flower varieties.