| 用于体质量估测的黄羽鸡姿态关键帧识别与分析 |
| |
| 引用本文： | 张小敏,徐涛,张延宁,高源,朱逸航,饶秀勤.用于体质量估测的黄羽鸡姿态关键帧识别与分析[J].农业机械学报,2022,53(12):254-263. |
| |
| 作者姓名： | 张小敏 徐涛 张延宁 高源 朱逸航 饶秀勤 |
| |
| 基金项目： | 浙江省重点研发计划项目（2021C02026） |
| |
| 摘 要： | 体质量是评价家禽生长状况的关键指标，但家禽姿态的变化会影响体质量估测精度。本研究提出了一种SE-ResNet18+fLoss网络对平养模式下黄羽鸡姿态关键帧进行识别，融合了注意力机制SE模块和残差结构，并改进了损失函数，通过Focal Loss监督信号来解决样本不平衡问题，同时引入梯度加权类激活图对末端分类规则的合理性进行解释。利用4295幅鸡只图像构建数据集，测试集中鸡只的站立、低头、展翅、梳理羽毛、坐姿和遮挡6类姿态情况识别的F1值分别为94.34%、91.98%、76.92%、93.75%、100%和93.68%；黄羽鸡姿态关键帧的识别精确率为97.38%、召回率为97.22%、F1值为97.26%、识别速度为19.84f/s，识别精度、召回率和F1值均优于ResNet18、MobileNet18 V2和SE-ResNet18网络，在提高黄羽鸡姿态关键帧识别精度的同时保证了实时性，为准确估测家禽体质量提供了技术支持。
|
| 关 键 词： | 黄羽鸡 姿态识别 体质量估测 SE-ResNet Focal Loss |
| 收稿时间： | 2022/8/1 0:00:00 |
Posture Key Frame Recognition and Analysis for Weight Estimation of Yellow-feathered Chickens |
| |
| ZHANG Xiaomin,XU Tao,ZHANG Yanning,GAO Yuan,ZHU Yihang,RAO Xiuqin.Posture Key Frame Recognition and Analysis for Weight Estimation of Yellow-feathered Chickens[J].Transactions of the Chinese Society of Agricultural Machinery,2022,53(12):254-263. |
| |
| Authors: | ZHANG Xiaomin XU Tao ZHANG Yanning GAO Yuan ZHU Yihang RAO Xiuqin |
| |
| Institution: | Zhejiang University |
| |
| Abstract: | Body weight is a key indicator to evaluate the growth condition of poultry. However, the variation of poultry posture will affect the accuracy of weight estimation. SE-ResNet18+fLoss network was proposed to detect the posture key frames of floor-reared yellow-feathered chickens. The attention mechanism SE module and residual structure were integrated. And the Focal Loss was added to solve the problem of sample imbalance. In addition, the Gradient-weighted Class Activation Mapping was introduced to explain the rationality of the end classification rule. The dataset was constructed by 4295 images of yellow-feathered chickens. The F1-score of the SE-ResNet18+fLoss model on the test set for the chicken situations recognition of six classes: standing, bowing head, spreading wing, grooming feather, sitting and occlusion were 94.34%, 91.98%, 76.92%, 93.75%, 100% and 93.68%, respectively. Towards the detection of key posture frames on chickens, the accuracy, recall, F1-score and detection speed were 97.38%, 97.22% 97.26% and 19.84f/s, respectively. And the detection accuracy, recall and F1-score were better than those of ResNet18, MobileNet V2 and SE-ResNet18 networks. The study ensured real-time performance while improving the accuracy of key posture frame recognition, which provided technical support for accurate estimation of poultry weight. |
| |
| Keywords: | Yellow-feathered chickens posture recognition weight estimation SE-ResNet Focal Loss |
|
| 点击此处可从《农业机械学报》浏览原始摘要信息 |
| 点击此处可从《农业机械学报》下载免费的PDF全文 |
|
