Meta-Album Medicinal Leaf: 1,000 Images of 25 Healthy Herb Species

A collection of 1,000 images of healthy leaves from 25 medicinal plant species, created for few-shot learning benchmarks. The dataset was derived from a larger source of approximately 1,800 images captured on a mobile phone and was preprocessed by Meta-Album in March 2022. Original data was released by authors S Roopashree and J Anitha in 2020.

Use Cases

• Benchmarking few-shot image classification models based on the 25 plant species classes.
• Training models for plant species identification using the standardized 128x128 pixel leaf images.
• Developing domain adaptation techniques for plant imagery, given the uniform white background mentioned in the description.
• Researching data-efficient learning methods in computer vision using a curated subset of the original medicinal leaf database.

Strengths

• Contains 1,000 images across 25 distinct classes, providing a structured testbed for few-shot learning.
• Images are preprocessed to a uniform 128x128 pixel resolution, ensuring consistency for model input.
• Released under a permissive CC BY 4.0 license, facilitating open use and redistribution.
• Part of the standardized Meta-Album benchmark suite, allowing for direct comparison with other image datasets.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is a curated mini version; the full scope and potential bias of the original 1,800-image collection are not detailed.

Provenance

Source

Medicinal Leaf Dataset on Mendeley Data, authored by S Roopashree and J Anitha.

Collection Method

Leaves were plucked from plants, placed on a white background, and captured with a mobile phone camera. Meta-Album cropped and resized the images.

Time Range

Original data released in October 2020; Meta-Album version created March 2022.

Freshness

Created for Meta-Album on 01 March 2022; last update date is unknown.