In the 5G environment, the errors of industrial robots in recognizing, detecting, and positioning objects are serious; in addition, the recognition speed and accuracy of industrial robots are unsatisfactory. Therefore, in order to solve these problems, the image layers are convolved and pooled through the deep learning model of artificial intelligence, and the visual recognition system of industrial robots are optimized through the advanced methods of the target classification algorithm. Then, the algorithm of the improved Fast R-CNN target detection model is verified by the bottled objects. The VGG-16 classification network based on Hyper-Column scheme is verified by targeting the bottled objects in small and complex environments. The results have shown that both the Fast R-CNN algorithm and the improved VGG-16 classification network based on the Hyper-Column scheme can complete the localization and recognition of the target objects, indicating that the Fast R-CNN algorithm and the Hyper-Column-based scheme have improved the accuracy and effectiveness of target recognition and positioning of VGG-16 classification network.

中文翻译：

---

在5G环境中使用机器视觉对工业机器人进行目标识别

在5G环境中，工业机器人在识别，检测和定位对象方面存在严重错误; 另外，工业机器人的识别速度和准确性也不令人满意。因此，为了解决这些问题，通过人工智能的深度学习模型对图像层进行卷积和合并，并通过目标分类算法的先进方法对工业机器人的视觉识别系统进行优化。然后，通过瓶装对象验证了改进后的快速R-CNN目标检测模型的算法。通过针对小型复杂环境中的瓶装对象，验证了基于超列方案的VGG-16分类网络。