Garden Flowers Canna Indica Identification Image Dataset

Currently, precise identification of plants poses a significant challenge in garden management and ecological monitoring. Traditional plant identification relies on human experience, leading to low efficiency and poor accuracy. Automatic identification technology based on computer vision is limited in its application due to the lack of high-quality training data. The aim of this dataset is to provide high-quality images of Canna indica for training identification algorithms, thereby improving the accuracy and efficiency of automatic garden plant identification. Data is collected using high-definition camera equipment, capturing multi-angle images under natural light conditions to ensure image clarity and diversity. Multiple rounds of expert annotation and consistency checks ensure the accuracy of data labels, with the annotation team comprising 10 professionals with a background in botany. The data undergoes preprocessing steps such as image denoising and size standardization, and is ultimately stored in JPG format, organized and arranged according to plant variety.