Maize Ear Sensing for On-Farm Yield Predictions
Nondestructive depth sensing that turns each captured husk-on maize ear into an instant, per-plant grain-yield estimate.
Pedro Cisdeli1 · Gustavo Nocera Santiago1 · German Mandrini1 · Ignacio Ciampitti1
1Purdue University - Dept. of Agronomy
01 — Abstract
The approach.
We introduce the first fully on-field pipeline that estimates maize-ear length, width and volume from a single RGB + depth capture and immediately forecasts grain yield per plant. A YOLOv12n-seg model isolates the ear in unconstrained lighting, a bespoke network (EVNet) regresses volume from the segmented point cloud, and gradient-boosted trees convert morphology into yield.
On Kansas field data we reach 98.6% mAP@0.5 for segmentation, 28.9 ml RMSE for volume, and 13.9 g RMSE for yield (ideal) / 24.1 g (real). The pipeline runs in approximately 1 s per image, needs no destructive sampling, and the images, code, and trained weights are open-sourced.
Real-time field deployment
02 — Method
Our Pipeline.
03 — Results
Key Numbers.
Segmentation
98.6%
mAP@0.5
95.8% precision / recall · 1.11s per image
Volume (Real-world)
R² 0.88
Goodness of fit
RMSE = 28.9 ml
Yield (Ideal)
R² 0.96
Goodness of fit
RMSE = 13.9 g
Yield (Real-world)
R² 0.89
Goodness of fit
RMSE = 24.1 g
04 — Impact
Why it matters.
- 01
First non-destructive ear yield predictor deployable in the field.
- 02
Open dataset & code (CornDepth) to accelerate follow-up work.
- 03
Bridges phenotyping & on-farm decision-making for breeders and agronomists.
05 — Resources
Paper, code & data.
Citation
@InProceedings{Cisdeli_2025_CVPR,
author = {Cisdeli, Pedro and Santiago, Gustavo Nocera and Mandrini, German and Ciampitti, Ignacio},
title = {Maize ear sensing for on-farm yield predictions},
booktitle = {Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR) Workshops},
month = {June},
year = {2025},
pages = {5402-5411}
}