Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt

The identification of individual cows is a prerequisite and foundation for realizing accurate and intelligent farming, but this identification method based on image information is easily affected by the environment and observation angle. To identify cows more accurately and efficiently, a novel indi...

Full description

Saved in:
Bibliographic Details
Main Authors: Kaixuan Zhao, Jinjin Wang, Yinan Chen, Junrui Sun, Ruihong Zhang
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Agriculture
Subjects:
Online Access:https://www.mdpi.com/2077-0472/15/7/710
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850184532878163968
author Kaixuan Zhao
Jinjin Wang
Yinan Chen
Junrui Sun
Ruihong Zhang
author_facet Kaixuan Zhao
Jinjin Wang
Yinan Chen
Junrui Sun
Ruihong Zhang
author_sort Kaixuan Zhao
collection DOAJ
description The identification of individual cows is a prerequisite and foundation for realizing accurate and intelligent farming, but this identification method based on image information is easily affected by the environment and observation angle. To identify cows more accurately and efficiently, a novel individual recognition method based on the using anchor point detection and body pattern features from top-view depth images of cows was proposed. First, the top-view RGBD images of cows were collected. The hook and pin bones of cows were coarsely located based on the improved PointNet++ neural network. Second, the curvature variations in the hook and pin bone regions were analyzed to accurately locate the hook and pin bones. Based on the spatial relationship between the hook and pin bones, the critical area was determined, and the key region was transformed from a point cloud to a two-dimensional body pattern image. Finally, body pattern image classification based on the improved ConvNeXt network model was performed for individual cow identification. A dataset comprising 7600 top-view images from 40 cows was created and partitioned into training, validation, and test subsets using a 7:2:1 proportion. The results revealed that the AP<sub>50</sub> value of the point cloud segmentation model is 95.5%, and the cow identification accuracy could reach 97.95%. The AP<sub>50</sub> metric of the enhanced PointNet++ neural network exceeded that of the original model by 3 percentage points. Relative to the original model, the enhanced ConvNeXt model achieved a 6.11 percentage point increase in classification precision. The method is robust to the position and angle of the cow in the top-view.
format Article
id doaj-art-520e024fe2bd41c9a4cc3408e6f1691f
institution OA Journals
issn 2077-0472
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Agriculture
spelling doaj-art-520e024fe2bd41c9a4cc3408e6f1691f2025-08-20T02:17:00ZengMDPI AGAgriculture2077-04722025-03-0115771010.3390/agriculture15070710Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXtKaixuan Zhao0Jinjin Wang1Yinan Chen2Junrui Sun3Ruihong Zhang4College of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471023, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471023, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471023, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471023, ChinaCollege of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471023, ChinaThe identification of individual cows is a prerequisite and foundation for realizing accurate and intelligent farming, but this identification method based on image information is easily affected by the environment and observation angle. To identify cows more accurately and efficiently, a novel individual recognition method based on the using anchor point detection and body pattern features from top-view depth images of cows was proposed. First, the top-view RGBD images of cows were collected. The hook and pin bones of cows were coarsely located based on the improved PointNet++ neural network. Second, the curvature variations in the hook and pin bone regions were analyzed to accurately locate the hook and pin bones. Based on the spatial relationship between the hook and pin bones, the critical area was determined, and the key region was transformed from a point cloud to a two-dimensional body pattern image. Finally, body pattern image classification based on the improved ConvNeXt network model was performed for individual cow identification. A dataset comprising 7600 top-view images from 40 cows was created and partitioned into training, validation, and test subsets using a 7:2:1 proportion. The results revealed that the AP<sub>50</sub> value of the point cloud segmentation model is 95.5%, and the cow identification accuracy could reach 97.95%. The AP<sub>50</sub> metric of the enhanced PointNet++ neural network exceeded that of the original model by 3 percentage points. Relative to the original model, the enhanced ConvNeXt model achieved a 6.11 percentage point increase in classification precision. The method is robust to the position and angle of the cow in the top-view.https://www.mdpi.com/2077-0472/15/7/710dairy cowsindividual identificationPointNet++body pattern featuresConvNeXt
spellingShingle Kaixuan Zhao
Jinjin Wang
Yinan Chen
Junrui Sun
Ruihong Zhang
Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
Agriculture
dairy cows
individual identification
PointNet++
body pattern features
ConvNeXt
title Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
title_full Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
title_fullStr Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
title_full_unstemmed Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
title_short Individual Identification of Holstein Cows from Top-View RGB and Depth Images Based on Improved PointNet++ and ConvNeXt
title_sort individual identification of holstein cows from top view rgb and depth images based on improved pointnet and convnext
topic dairy cows
individual identification
PointNet++
body pattern features
ConvNeXt
url https://www.mdpi.com/2077-0472/15/7/710
work_keys_str_mv AT kaixuanzhao individualidentificationofholsteincowsfromtopviewrgbanddepthimagesbasedonimprovedpointnetandconvnext
AT jinjinwang individualidentificationofholsteincowsfromtopviewrgbanddepthimagesbasedonimprovedpointnetandconvnext
AT yinanchen individualidentificationofholsteincowsfromtopviewrgbanddepthimagesbasedonimprovedpointnetandconvnext
AT junruisun individualidentificationofholsteincowsfromtopviewrgbanddepthimagesbasedonimprovedpointnetandconvnext
AT ruihongzhang individualidentificationofholsteincowsfromtopviewrgbanddepthimagesbasedonimprovedpointnetandconvnext