Strawberry 3D Point Clouds for Shape Completion and Weight Estimation

Pengyu Li's dataset contains 400 individual strawberry fruit 3D point clouds. It supports computer vision tasks like single-view shape completion, shape classification, and weight estimation. The dataset was last updated in March 2026.

Use Cases

• Train single-view shape completion models using partial point clouds to predict full 3D strawberry shapes.
• Classify strawberry shapes (e.g., conical, elongated) directly from point cloud data features.
• Develop regression models to estimate strawberry weight from 3D point cloud features like volume or surface area.
• Benchmark point cloud processing networks for agricultural object reconstruction tasks.

Strengths

• Contains 400 individual strawberry samples, providing a substantial collection for model training.
• Data is specifically formatted for three distinct tasks: shape completion, classification, and weight estimation.
• Released under a permissive CC-BY-4.0 license, allowing for wide reuse and modification.

Limitations

• The dataset size of 400 samples may be limited for training large, deep learning models without augmentation.
• Geographic origin and cultivation conditions of the strawberries are not specified, which could limit generalizability.
• No information is provided on potential class imbalance across different strawberry shapes or sizes.

Provenance

Source

Pengyu Li via figshare.

Collection Method

3D scanning of individual strawberry fruits to create point clouds.

Freshness

Last updated March 2026.