LPO-YOLOv5s: A Lightweight Pouring Robot Object Detection Algorithm

Abstract

The casting process involves pouring molten metal into a mold cavity. Currently, traditional object detection algorithms exhibit a low accuracy and are rarely used. An object detection model based on deep learning requires a large amount of memory and poses challenges in the deployment and resource allocation for resource limited pouring robots. To address the accurate identification and localization of pouring holes with limited resources, this paper designs a lightweight pouring robot hole detection algorithm named LPO-YOLOv5s, based on YOLOv5s. First, the MobileNetv3 network is introduced as a feature extraction network, to reduce model complexity and the number of parameters. Second, a depthwise separable information fusion module (DSIFM) is designed, and a lightweight operator called CARAFE is employed for feature upsampling, to enhance the feature extraction capability of the network. Finally, a dynamic head (DyHead) is adopted during the network prediction stage, to improve the detection performance. Extensive experiments were conducted on a pouring hole dataset, to evaluate the proposed method. Compared to YOLOv5s, our LPO-YOLOv5s algorithm reduces the parameter size by 45% and decreases computational costs by 55%, while sacrificing only 0.1% of mean average precision (mAP). The model size is only 7.74 MB, fulfilling the deployment requirements for pouring robots.

YOLOv5-based Student Counting Software Design

“…The counting software in this study is realized through Qt Designer, the basic idea is: design the interface of the counting software in Qt Designer, including the specific keys to be used and the display interface, after designing the software, save it as a .ui file, because this model is realized through python code, so by executing the code (1) at the command line to complete the .ui file to .py pyuic5 input.ui -o output.py (1) Where input.ui is the designed interface file and output.py is the output python file, after that, you only need to write the connection between the signal and each slot in the generated. py file [9], associate the events or actions on the interface with the back-end logic, and write the corresponding code to handle the interface operation and the connection logic of the training model. The perfect combination of interface and functionality can be realized.…”

Design, simulation, control of a hybrid pouring robot: enhancing automation level in the foundry industry

“…In contrast, deep learning-based control algorithms have been widely adopted in various robotics applications due to their higher accuracy and faster speed, which can effectively improve the control accuracy and fast model prediction [19]- [20]. Tianze et al adopted recurrent neural networks to do the experiment of the dynamic model predictive control of the water and obtained the optimal speed for precise dumping [22].…”