Improved P2PNet with Multi-Scale Feature Fusion and Point-Relation Refinement for Image-Based Maize Kernel Counting

Accurate counting of visible maize kernels from RGB images is important for maize phenotyping and yield-related trait assessment, but remains challenging because kernels are densely distributed and frequently affected by occlusion and local appearance similarity. Point-based counting provides a promising alternative to bounding-box-based detection by reducing annotation requirements while retaining object-level localization. However, the original P2PNet relies on a single prediction feature with limited direct interaction between high-resolution spatial information and high-level semantic features. To address this limitation, a bidirectional feature pyramid network (BiFPN) was incorporated into P2PNet to enable bidirectional multi-scale feature fusion among C3, C4, and C5 while retaining P4 as the prediction level. A Point Relation Refinement Module (PRM) was further investigated to examine the effects of candidate-point relational modeling on initial predictions through multi-head self-attention and residual refinement. Experiments were conducted on a maize kernel dataset containing 452 RGB images and 92,314 point annotations. Across three random seeds, the combined BiFPN+PRM model achieved the lowest mean absolute error (MAE) of 26.33±1.78 and mean relative error (MRE) of 10.50±0.48% among the four main P2PNet ablation configurations, compared with 27.70±1.92 and 11.54±0.83% for the original P2PNet. BiFPN alone did not consistently improve counting accuracy but reduced computational complexity from 117.474 to 105.159 GFLOPs and increased inference throughput from 279.98 to 288.90 FPS. PRM exhibited sample-dependent behavior and substantial computational overhead. However, top-512 candidate selection reduced inference latency by 74.44% compared with global attention under the actual-resolution evaluation protocol, with slightly lower mean MAE and RMSE but a higher mean MRE (10.50% to 11.01%). These findings demonstrate the potential of the proposed framework for non-destructive image-based kernel quantification in maize phenotyping.

Authors

Institutions

Publication Details

Journal
AgriEngineering
Published
2026-11-12
DOI
https://doi.org/10.3390/agriengineering8110487
Primary Topic
Smart Agriculture and AI
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Improved P2PNet with Multi-Scale Feature Fusion and Point-Relation Refinement for Image-Based Maize Kernel Counting

Gaopeng Zhang, Yingchun Kuang, Renfeng Zhang, Xiaotao Fu
AgriEngineering
Smart Agriculture and AI
article

Improved P2PNet with Multi-Scale Feature Fusion and Point-Relation Refinement for Image-Based Maize Kernel Counting

Gaopeng Zhang, Yingchun Kuang, Renfeng Zhang, Xiaotao Fu
article en

Abstract

Accurate counting of visible maize kernels from RGB images is important for maize phenotyping and yield-related trait assessment, but remains challenging because kernels are densely distributed and frequently affected by occlusion and local appearance similarity. Point-based counting provides a promising alternative to bounding-box-based detection by reducing annotation requirements while retaining object-level localization. However, the original P2PNet relies on a single prediction feature with limited direct interaction between high-resolution spatial information and high-level semantic features. To address this limitation, a bidirectional feature pyramid network (BiFPN) was incorporated into P2PNet to enable bidirectional multi-scale feature fusion among C3, C4, and C5 while retaining P4 as the prediction level. A Point Relation Refinement Module (PRM) was further investigated to examine the effects of candidate-point relational modeling on initial predictions through multi-head self-attention and residual refinement. Experiments were conducted on a maize kernel dataset containing 452 RGB images and 92,314 point annotations. Across three random seeds, the combined BiFPN+PRM model achieved the lowest mean absolute error (MAE) of 26.33±1.78 and mean relative error (MRE) of 10.50±0.48% among the four main P2PNet ablation configurations, compared with 27.70±1.92 and 11.54±0.83% for the original P2PNet. BiFPN alone did not consistently improve counting accuracy but reduced computational complexity from 117.474 to 105.159 GFLOPs and increased inference throughput from 279.98 to 288.90 FPS. PRM exhibited sample-dependent behavior and substantial computational overhead. However, top-512 candidate selection reduced inference latency by 74.44% compared with global attention under the actual-resolution evaluation protocol, with slightly lower mean MAE and RMSE but a higher mean MRE (10.50% to 11.01%). These findings demonstrate the potential of the proposed framework for non-destructive image-based kernel quantification in maize phenotyping.

AgriEngineeringVol. 8(11)
Hunan Agricultural University (CN)
Hunan Agricultural University
Zero hunger
Smart Agriculture and AI
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon