Amaryllis Image Recognition Dataset

Currently, with the rapid development of the horticulture industry, the diversification and increase in the number of flower varieties pose challenges to flower management. Current solutions such as manual recognition and recording are inefficient and prone to errors. The establishment of the Amaryllis Image Recognition Dataset aims to leverage advanced image recognition technology to achieve fast and accurate flower classification, optimizing garden management processes. Data collection is conducted using high-definition cameras under natural light to ensure the capture of complete plant features. Data quality is ensured through multiple rounds of annotation and consistency checks, performed by a team of experts in the field of botany, comprising more than 10 members. After annotation, the data undergo image enhancement and denoising processing and are finally stored in JPG format, organized in a categorized directory for easy access and model training.