Research on Chengdu Ma Goat Recognition Based on Computer Vison

SIMPLE SUMMARY: Accurate and efficient automatic individual recognition of livestock such as goats is one of the key tasks to achieve smart farming. However, this task is challenging due to factors such as the dense distribution of livestock in a barn. To accurately recognize Chengdu ma goats, a high-quality sheep breed with local characteristics, we construct a Chengdu ma goat dataset containing nearly one thousand images with manual annotation information and propose a method for the automatic individual recognition of Chengdu ma goats using a computer. Experiments demonstrate the usability of our method to accurately recognize Chengdu ma goats under extreme conditions such as dense target distribution and drastic differences in scale. It is worth noting that the method is scalable, which enhances the network’s representation ability for data by learning the structural information of goats’ head area. The location and category information obtained by our automatic individual recognition method is the basis for realizing a series of subsequent intelligent breeding tasks, such as precision feeding based on age and sex. Our method provides a new idea for automatic livestock recognition and can be extended to be applied to additional types of livestock. ABSTRACT: The Chengdu ma goat is an excellent local breed in China. As one of the breeds listed in the National List of Livestock and Poultry Genetic Resources Protection, the protection of its germplasm resources is particularly important. However, the existing breeding and protection methods for them are relatively simple, due to the weak technical force and lack of intelligent means to assist. Most livestock farmers still conduct small-scale breeding in primitive ways, which is not conducive to the breeding and protection of Chengdu ma goats. In this paper, an automatic individual recognition method for Chengdu ma goats is proposed, which saves labor costs and does not depend on large-scale mechanized facilities. The main contributions of our work are as follows: (1) a new Chengdu ma goat dataset is built, which forms the basis for object detection and classification tasks; (2) an improved detection algorithm for Chengdu ma goats based on TPH-YOLOv5 is proposed, which is able to accurately localize goats in high-density scenes with severe scale variance of targets; (3) a classifier incorporating a self-supervised learning module is implemented to improve the classification performance without increasing the labeled data and inference computation overhead. Experiments show that our method is able to accurately recognize Chengdu ma goats in the actual indoor barn breeding environment, which lays the foundation for precision feeding based on sex and age.