Carrot Appearance Inspection Dataset

The current agricultural industry faces challenges such as low crop quality inspection efficiency and high labor costs. Existing inspection methods largely rely on manual visual checks, which are subjective and inefficient. This dataset aims to solve the data scarcity problem in object detection tasks by providing high-quality carrot image data, thereby enhancing the level of automation in smart agriculture. Data collection is conducted using high-resolution cameras in well-lit greenhouse environments to ensure the clarity and recognizability of each image. In terms of quality control, multiple rounds of labeling and expert reviews are employed to ensure labeling accuracy and consistency. All data is stored in JPG format, organized into separate folders by category and quality for ease of subsequent use and retrieval.