Welding defects on new energy batteries based on 2D pre-processing and improved-region-growth method in the small field of view

Abstract

Abstract The assessment of welding quality in battery shell production is a crucial aspect of battery production. Battery surface reconstruction can inspect the quality of the weld instead of relying on human inspection. This paper proposes a defect detection method in the small field of view based on 2D pre-processing and an improved-region-growth method. A novel approximation-based, high-precision, and simple operation method for line structure optical plane calibration under small field of view is presented, with a measurement error within 0.01 mm. By pre-processing the line scan 2D images, the defect location distribution is obtained, and then the images near the abnormal points are reconstructed in third dimensional (3D). The proposed method enables the extraction of the morphology, size, and other information of the defects with high accuracy. The results of various defect detection experiments demonstrate the stable and reliable performance of the system. The experimental results of defect recognition rate are over 95.3% for defects above 0.5 mm in diameter, and the inspection time is less than 1/2 of the direct 3D defect inspection. Overall, this method has proven to be highly effective in the assessment of welding quality in new energy battery production.