| 基于改进YOLO v5s的经产母猪发情检测方法研究 |
| |
| 引用本文： | 薛鸿翔,沈明霞,刘龙申,陈金鑫,单武鹏,孙玉文.基于改进YOLO v5s的经产母猪发情检测方法研究[J].农业机械学报,2023,54(1):263-270. |
| |
| 作者姓名： | 薛鸿翔 沈明霞 刘龙申 陈金鑫 单武鹏 孙玉文 |
| |
| 作者单位： | 南京农业大学 |
| |
| 基金项目： | 江苏省科技计划项目（BE2021363） |
| |
| 摘 要： | 为解决限位栏场景下经产母猪查情难度大、过于依赖公猪试情和人工查情的问题，提出了一种基于改进YOLO v5s算法的经产母猪发情快速检测方法。首先，利用马赛克增强方式(Mosaic data augmentation, MDA)扩充数据集，以丰富数据表征；然后，利用稀疏训练(Sparse training, ST)、迭代通道剪枝(Network pruning, NP)、模型微调(Fine tune, FT)等方式重构模型，实现模型压缩与加速；最后，使用DIOU＿NMS代替GIOU＿NMS,以提高目标框的识别精度，确保模型轻量化后，仍保持较高的检测精度。试验表明，优化后的算法识别平均精确率可达97.8%,单幅图像平均检测时间仅1.7 ms,单帧视频平均检测时间仅6 ms。分析空怀期母猪发情期与非发情期的交互行为特征，发现母猪发情期较非发情期交互时长与频率均显著提高(P<0.001)。以20 s作为发情检测阈值时，发情检测特异性为89.1%、准确率为89.6%、灵敏度为90.0%,该方法能够实现发情母猪快速检测。
|
| 关 键 词： | 经产母猪 发情检测 深度学习 改进YOLO v5s |
| 收稿时间： | 2022/3/22 0:00:00 |
Estrus Detection Method of Parturient Sows Based on Improved YOLO v5s |
| |
| XUE Hongxiang,SHEN Mingxi,LIU Longshen,CHEN Jinxin,SHAN Wupeng,SUN Yuwen.Estrus Detection Method of Parturient Sows Based on Improved YOLO v5s[J].Transactions of the Chinese Society of Agricultural Machinery,2023,54(1):263-270. |
| |
| Authors: | XUE Hongxiang SHEN Mingxi LIU Longshen CHEN Jinxin SHAN Wupeng SUN Yuwen |
| |
| Institution: | Nanjing Agricultural University |
| |
| Abstract: | Quickly and accurately identify estrous sows to ensure timely breeding is the key to keep sow breeding performance. Aiming at the problems of low sensitivity and accuracy rate in sows estrus identification, according to the interactive characteristics of sows and bionic boars, an estrus detection method of sows based on improved YOLO v5s was proposed. Firstly, the automatic inspection robot was used to collect the video data of sows estrus behavior. The Mosaic data augmentation was used to expand the data set to enrich the data representation and enhance the robustness of the detection model, and the estrus detection model based on YOLO v5s was constructed and then optimized by sparse training, iterative channel pruning and fine-tuning to realize model compression and acceleration. DIOU_NMS was used to replace the GIOU_NMS to improve the recognition accuracy and keep high detection accuracy with lightweight-model. The results showed that the average accuracy of the algorithm was 97.8%, the average detection time of each picture was 1.7ms, and the average detection time of each video frame was 6ms. Analyzing the interactive behavior characteristics of estrous and non-estrous sows at the end of lactation, it was found that the interactive duration and frequency of estrous sows were significantly higher than that of non-estrous sows (P<0.001). On this basis, it was found that when 20s was used as the threshold of estrus detection, the sensitivity rate of estrus detection was 90.0%, the accuracy rate was 89.6%, and the specificity was 89.1%, the method can be used to rapidly and accurately detect estrus sows. |
| |
| Keywords: | parturient sows estrus detection deep learning improved YOLO v5s |
|
| 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 |
|
