Even with the introduction of target detection technology, it remains challenging to efficiently count eggs in chicken coops due to the many interfering factors present in real farming environments, such as high density, shading, and nearby scenery. Moreover, target detection lacks association with temporal information, resulting in frequent disappearance of bounding boxes around eggs. Additionally, the limited space in chicken coops makes it unsuitable to deploy models on large servers. Therefore, the development of a reliable and lightweight egg counting algorithm applicable for edge devices is necessary in on-site chicken coops.

翻译以下内容为英文要求英语地道、学术化。即使引入目标检测技术还是很难在鸡舍现场高效地数数鸡蛋。原因在于真实养殖场包含较多干扰因素如高密度、遮荫或近景色干扰而目标检测缺乏对时序信息的关联导致鸡蛋的包围盒会频繁消失并且鸡舍现场空间有限不适合将模型部署在体积较大服务器上。因此在鸡舍现场开发可靠、轻量适用于边缘设备部署的鸡蛋数数算法是必要的。

原文地址: https://www.cveoy.top/t/topic/bSyY 著作权归作者所有。请勿转载和采集!

免费AI点我,无需注册和登录