Abstract:Energy storage devices play a crucial role in balancing energy supply and enhancing energy utilization efficiency. Lithium batteries have become one of the mainstream choices for energy storage devices due to their high energy density and long cycle life. However, the manufacturing process of lithium batteries faces numerous quality control challenges, particularly with defects in battery electrodes. In this paper, an improved detection model, which includes edge feature enhancement module and multi-scale feature extraction module, is designed to solve the four major defects in the production of battery electrode pieces, namely scratches, clumps, cracks and bubbles. Multiple convolution kernel cascades are used for feature fusion to improve the extraction capability of multi-scale defect targets. Experimental results show that the model demonstrates high accuracy, increasing the overall defect recognition rate from 88.9% to 95.9%, and accurately identifies and locates defects occurring during battery electrode production. This research provides an efficient solution for defect detection in lithium battery electrodes, promoting quality improvement and safety assurance in lithium battery energy storage devices.