An Integrated Goat Head Detection and Automatic Counting Method Based on Deep Learning.

Abstract

Simple SummaryTo achieve precision and intelligence in farming, automatic recognition and counting of goats are important and fundamental parts of the process of large-scale goat farming. Currently, many farms with low modernization use manual counting, which has the obvious shortcomings of low efficiency and difficulty in avoiding duplication and omissions due to the large population base and frequent counting needs of goats. In order to solve this problem in the farming process, an efficient and accurate goat counting method is urgently needed. In this study, we address the above problem by constructing an integrated deep learning model for automatic detection and counting of goats based on computer vision technology with the Chengdu Ma goat as the research object. It is worth noting that we have improved the model using a series of advanced and effective strategies to enhance the performance of the model. Experiments show that our method can achieve accurate automatic counting of goats in a practical breeding environment. The method is beneficial to the regionalized management of goat barns and can be applied to different goat species with high practicality.Goat farming is one of the pillar industries for sustainable development of national economies in some countries and plays an active role in social and economic development. In order to realize the precision and intelligence of goat breeding, this paper describes an integrated goat detection and counting method based on deep learning. First, we constructed a new dataset of video images of goats for the object tracking task. Then, we took YOLOv5 as the baseline of the object detector and improved it using a series of advanced methods, including: using RandAugment to explore suitable data augmentation strategies in a real goat barn environment, using AF-FPN to improve the network’s ability to represent multi-scale objects, and using the Dynamic Head framework to unify the attention mechanism with the detector’s heads to improve its performance. The improved detector achieved 92.19% mAP, a significant improvement compared to the 84.26% mAP of the original YOLOv5. In addition, we also input the information obtained by the detector into DeepSORT for goat tracking and counting. The average overlap rate of our proposed method is 89.69%, which is significantly higher than the 82.78% of the original combination of YOLOv5 and DeepSORT. In order to avoid double counting as much as possible, goats were counted using the single-line counting based on the results of goat head tracking, which can support practical applications.