A recent project captured a strawberry from 90 angles, producing 88 focus‑stacked images. The shots were taken with a Nikon Z8 full‑frame camera, f/7.1, 1/160 s, ISO 100, and a Laowa 180mm macro lens. LED lighting and a bluescreen background helped isolate the fruit for clean composites in a controlled studio setting to ensure consistent lighting effortlessly.

The dataset behind the images is available under a permissive CC BY license, encouraging reuse. The creator also released a training pipeline named slang‑splat, hosted on GitHub. This tool leverages the 90‑perspective set to generate a high‑resolution 3D reconstruction of the fruit, demonstrating how multi‑view data can feed neural surface models in real world applications.

Users interested in the raw assets can download them from the author's Patreon, where the COLMAP dataset is also offered free of charge. The images can be combined with the slang‑splat code to produce photorealistic renderings or to train machine‑learning models for object recognition and depth estimation in high detail projects across educational and industrial.

By exposing a complete workflow—from capture through processing to licensing—this initiative lowers the barrier for researchers and hobbyists alike. The explicit focus stacking, open‑source tools, and generous sharing model exemplify a trend toward democratizing photogrammetry and AI‑driven visual content creation that enables rapid prototype development and creative exploration within the visual engineering community and beyond.