This system is mainly used to detect the new energy battery case with high precision. It can be used for high-speed concurrent detection, automatic defect measurement and automatic classification. It adopts hardware+software mode to detect battery case defects.

This equipment utilizes machine vision to replace the human eye for conducting comprehensive inspections on all six sides of the power battery's appearance. The inspections cover a wide range of defects, including cracks, damages, missing components, missing installation, contamination, wrong installation, recognition of QR codes or characters on the battery pack, and skewing of PIN pins, among others, thereby ensuring the safety performance of batteries.

The equipment is used to inspect the welding quality of connecting pieces, and can effectively inspect defects such as solder skips, incomplete welding tracks, weld deviations, cold welds, weld holes gaps, explosions, bubbles, arc pits, etc,. The equipment scans each connecting piece position with a 3D camera and outputs the brightness map and depth map of each point. For the brightness map captured by the 3D camera, the traditional algorithm is first used to extract the welding area of the connecting piece, and then the semantic segmentation algorithm is used to classify each pixel in the welding area to extract its welding tracks, poles, pits, cold weld, tracks explosion, bubbles, arc pits and other regional ROI. Find the corresponding depth information for combined processing, and finally match it with the standard value to remove it.