A Horse Facial Individual Recognition Algorithm Based on Improved YOLOv11n
Year: 2026
Type: article
Abstract: Rapid and accurate horse identity identification is an important prerequisite for the digital and intelligent management of livestock breeding. Traditional horse identification methods struggle to balance inference speed and recognition accuracy. In view of this, this paper takes a self-built dataset as the research object and proposes a horse identity detection algorithm based on YOLOv11n. The algorithm mainly makes three improvements: first, the traditional Conv layer is replaced by the Adaptive Kernel Convolution (AKConv) layer to enhance the adaptability of the model to targets of different sizes; second, an improved bi-level routing attention mechanism (BiFormer) is introduced into the C2PSA module to form the C2PSA-BiFormer module, which can efficiently capture long-range dependencies and effectively improve the performance of the model; finally, based on the idea of the Inner-IoU auxiliary box, the CIoU loss function is improved to further enhance the detection accuracy of the model. The experimental results show that, compared with the original model, the precision, recall, and mAP are increased by 7.5, 9.4, and 3.1 percentage points, respectively. The algorithm can effectively realize rapid horse identification and provides a methodological reference for the digitalization and intelligence of livestock breeding, showing high application value.
Source: Sensors
DOI: 10.3390/s26227317
Authors: Guzhalayi Bali(Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft), Xiaojia Zuo(Conceptualization, Investigation, Methodology, Validation, Writing – original draft), Xinwen Chen(Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing), Wenxin Zheng(Resources, Writing – review & editing), Saidiguli Samaiti(Data curation, Investigation, Resources)
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
SDG: Zero hunger
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7236719571
