Peruvian Amazon Tree Images for Species Classification

Encompassing 25,000 macroscopic images of trees from the Peruvian Amazon, representing 25 distinct species. The images were collected and validated by a botanical team using mobile device cameras and digital microscopes, with each image having a resolution of 480x640 pixels and three color channels.

Use Cases

• Train a convolutional neural network on 25,000 macroscopic images to classify the 25 Peruvian Amazon tree species.
• Build a multi-modal classifier using images captured by both mobile device cameras and digital microscopes.
• Analyze image features from the 480x640 pixel, three-channel images for species differentiation.
• Validate automated labeling systems against the manually validated species labels provided by botanical experts.

Strengths

• Contains 25,000 manually validated images, providing a substantial sample for model training.
• Covers 25 distinct tree species from a specific biogeographic region, the Peruvian Amazon.
• Images have consistent technical specifications: 480x640 pixel resolution with three color channels.
• Data collection employed two capture methods: mobile device cameras and digital microscopes.

Limitations

• The dataset is limited to 25 species from a single geographic region (Peruvian Amazon), which may not generalize to other forests.
• Image resolution is fixed at 480x640 pixels, which is lower than modern high-definition standards and may limit fine-grained feature extraction.
• The dataset's size and diversity are constrained by the manual collection and validation process of the botanical team.

Provenance

Source

Maderapp's botanical team, authored by anvelezec on Hugging Face.

Collection Method

Images were collected, validated, and labeled manually by a botanical team using mobile device cameras and digital microscopes.

Freshness

Last updated on 2022-05-25.

Geography

Peruvian Amazonia.