Search

The control of quality is an obligatory stage of any welding work which makes problem of timely detection of welding defects is important for industry nowadays. Through digitisation it possible to solve this problem more efficiently and eliminate the human factor. This paper presents a real-time industrial automated video analytics system for weld defect detection. The system uses deep learning to analyze video images and detect weld defects in real-time. Experiments showed that the system can effectively detect weld defects with high accuracy and speed. A safety and trustworthiness analysis of the system showed that it can be reliable and safe for use in industrial manufacturing. The proposed solution has several key advantages for the industry: it allows for real-time data processinga and requires low computing power, making it an energy-efficient and cost-effective solution.

Video Surveillance Systems (VSS) are progressing towards the industry digitalization. In this demo, we apply the Artificial Intelligence (AI) technology to detection of key points for a rainbow trout. The well-known YOLO-Pose neural network is typically used for human pose recognition in video recordings. We use this neural network for detection of 8 key point for an individual fish. Our constructed dataset includes 10 500 images for approximately 30 thousand fishes. The YOLO-Pose network is trained on the dataset using our combined loss function. Our early experiment shows that that detection works fine even in muddy underwater conditions.