Research on the Vision-Based Dairy Cow Ear Tag Recognition Method

Tianhong Gao,Daoerji Fan,Huijuan Wu,Xiangzhong Chen,Shihao Song,Yuxin Sun,Jia Tian

DOI: https://doi.org/10.3390/s24072194

IF: 3.9

2024-03-30

Sensors

Abstract:With the increase in the scale of breeding at modern pastures, the management of dairy cows has become much more challenging, and individual recognition is the key to the implementation of precision farming. Based on the need for low-cost and accurate herd management and for non-stressful and non-invasive individual recognition, we propose a vision-based automatic recognition method for dairy cow ear tags. Firstly, for the detection of cow ear tags, the lightweight Small-YOLOV5s is proposed, and then a differentiable binarization network (DBNet) combined with a convolutional recurrent neural network (CRNN) is used to achieve the recognition of the numbers on ear tags. The experimental results demonstrated notable improvements: Compared to those of YOLOV5s, Small-YOLOV5s enhanced recall by 1.5%, increased the mean average precision by 0.9%, reduced the number of model parameters by 5,447,802, and enhanced the average prediction speed for a single image by 0.5 ms. The final accuracy of the ear tag number recognition was an impressive 92.1%. Moreover, this study introduces two standardized experimental datasets specifically designed for the ear tag detection and recognition of dairy cows. These datasets will be made freely available to researchers in the global dairy cattle community with the intention of fostering intelligent advancements in the breeding industry.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The paper mainly studies a vision-based method for identifying cow ear tags to solve the challenges of individual cow recognition in modern farms. In large-scale farming, accurate management and non-invasive individual identification are crucial, and ear tags are the unique identifiers for identifying cows. Current methods include manual identification (time-consuming and inefficient) and automatic identification based on deep learning and traditional image processing techniques (with difficulties such as small object detection and complex scene recognition).

The main innovations of the paper include:

1. Proposing a lightweight Small-YOLOV5s model specifically designed for small object detection. It improves detection accuracy and speed, reduces the number of parameters and computational cost, and is suitable for real-time applications. It can also adapt to various interferences such as lighting conditions, tilting, and occlusion.

2. Using DBNet combined with Convolutional Recurrent Neural Network (CRNN) to achieve fast and accurate recognition of ear tag numbers. This approach avoids preprocessing operations such as character-level segmentation and image orientation correction. During the training stage, a mixture of handwritten and printed digital images is used to improve the model's recognition accuracy for both fonts.

3. Creating two standardized experimental datasets for ear tag detection and recognition. These datasets are freely available for global dairy industry researchers to promote the intelligent development of the breeding industry.

Through these methods, the paper addresses the accuracy and real-time issues of cow ear tag detection and recognition, providing technical support for precise management in farms.