GE 531X121PCRALG1
由于船体零部件种类繁多,CCD识别模板开发工作量大,本方案还提供了基于深度学习的识别解决方案。只需为每类零部件拍摄少量照片,识别软件即可自动学习到此类工件的特征,在实际应用中准确判断工件的类别,达到分拣的目的。相较于传统的人工识别,极大地提高了效率,减少误判。
GE 531X121PCRALG1
At the same time, the solution software system can intelligently plan the grabbing sequence based on the type of workpiece, the position of placement, and the type of gripper required for the encoded recognition of the workpiece, achieving the current gripper sorting to complete all the workpieces as much as possible, reducing the number of gripper replacements, and improving sorting efficiency.
Fusion deep learning
Quickly identify and sort workpieces
Due to the wide variety of ship components and the heavy workload of developing CCD recognition templates, this solution also provides a deep learning based recognition solution. By taking a small number of photos for each type of component, the recognition software can automatically learn the characteristics of such workpieces, accurately determine the category of workpieces in practical applications, and achieve the purpose of sorting. Compared to traditional manual recognition, it greatly improves efficiency and reduces false positives.
咨询:GE 531X121PCRALG1