Computer Vision-Based Egg Weight Measurement Improves Poultry Farm Efficiency

Strategic Insights -

Researchers at Zhejiang University have proposed a method for precise egg weight measurement in laying hen farms using a computer vision-based approach. This method integrates two artificial neural networks, Central differential-EfficientViT YOLO (CEV-YOLO) and Egg Weight Measurement Network (EWM-Net), to accurately segment eggs and estimate their weights. The study found that CEV-YOLO outperforms other YOLO-based models in egg segmentation, achieving a precision of 98.9%, a recall of 97.5%, and an Average Precision (AP) at an Intersection over Union (IoU) threshold of 0.9 (AP90) of 89.8%. The proposed pipeline also exhibits improved performance in practical production scenarios, with an R[superscript]2 of 0.926 and a mean absolute error (MAE) of 0.88 g in egg weight measurement.

Key Takeaways:

  • The researchers have proposed a computer vision-based method for precise egg weight measurement in laying hen farms, which integrates two artificial neural networks: CEV-YOLO and EWM-Net.
  • CEV-YOLO outperforms other YOLO-based models in egg segmentation, achieving a precision of 98.9%, a recall of 97.5%, and an Average Precision (AP) at an Intersection over Union (IoU) threshold of 0.9 (AP90) of 89.8%.
  • The proposed pipeline exhibits improved performance in practical production scenarios, with an R[superscript]2 of 0.926 and a mean absolute error (MAE) of 0.88 g in egg weight measurement.
  • The study aims to improve the accuracy and efficiency of feed-to-egg ratio measurement in laying hen farms.
  • The proposed method has the potential to deploy in practical production scenarios to improve egg weight measurement precision.

Statistics:

  • Precision of CEV-YOLO in egg segmentation: 98.9%
  • Recall of CEV-YOLO in egg segmentation: 97.5%
  • Average Precision (AP) of CEV-YOLO at an Intersection over Union (IoU) threshold of 0.9 (AP90): 89.8%
  • R[superscript]2 of EWM-Net in egg weight measurement: 0.926
  • Mean Absolute Error (MAE) of EWM-Net in egg weight measurement: 0.88 g

Sources:

  • Agriculture, 2025,15(19):2035. (Agriculture - http://www.mdpi.com/journal/agriculture)
  • NewsRx. Zhejiang University Researchers Highlight Research in Agriculture (Computer Vision-Based Multi-Feature Extraction and Regression for Precise Egg Weight Measurement in Laying Hen Farms). Computer Weekly News. October 29, 2025; p 935.