HAE-Net: A hierarchy-aware enhanced method for 3D panoptic segmentation in orchard scenes
Shuo Zhang, Yiqun Wang, Lianchuan Shi, Ze Ji, Xiude Chen, Weikuan Jia · 2026
Read HAE-Net: A hierarchy-aware enhanced method for 3D panoptic segmentation in orchard scenes by Shuo Zhang, Yiqun Wang, Lianchuan Shi, Ze Ji, Xiude Chen, Weikuan Jia (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13634764.txt.
Description
Shuo Zhang and five co-authors present a computer vision method for making sense of three-dimensional scans of orchards, in this 2026 paper. Robots that pick fruit or dose fertilizer need to find each fruit, count it, and know which tree it hangs on. Laser scans give the shape of a scene without depending on daylight, but software must sort the points into objects at two levels at once: single fruits and trunks, and whole trees.
Their system, HAE-Net, uses one shared reader of the scan feeding three decoders, one for each level of the task. Added parts sharpen edges between touching fruits and group them under their parent tree. Most earlier methods work at only one level and miss that family link. A tool that keeps both levels straight could sharpen yield forecasts and help machines plan harvests tree by tree.
Catalog record
| CREATOR | Shuo Zhang, Yiqun Wang, Lianchuan Shi, Ze Ji, Xiude Chen, Weikuan Jia |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |