Intelligent 3D Visual defect Detection for Industrial Applications

Abstract Currently, new energy industries such as electric vehicles and storage batteries are experiencing rapid growth throughout the world. As a recognized ideal element, lithium have also attracted considerable interest. To improve image contrast, industry employs preprocessing algorithms. Due to high requirement of consistency in template matching method, which is difficult meet practice, large apparent difference between defects, it challenging design features. Therefore, framework for defect detection based on method deep learning with strong feature expression capability. In light severe imbalance number samples detection, data augmentation generation methods used simulate real order training effect neural networks alleviate burden annotation some extent.