Abstract: With the rapid advancement of solar technology, efficient methods for detecting defects in solar cell panels have become crucial. Traditional defect detection methods often fail to meet the requirements of industrial applications because of their inadequate precision, particularly in detecting minute or subtle defects. To address this issue, this paper presents an enhanced YOLOv8n algorithm that introduces new structures and a loss function to improve the detection effectiveness. Initially, the study designed a C2f_RVB module that employs RepViTBlock technology to optimize the feature expression capability and effectively reduce the model parameters. By enhancing the expression of deep features, the C2f_RVB module significantly improves the accuracy of detecting small-scale defects. Additionally, the introduced TFE module enhances the capability of the model to detect minute defects through improved multiscale feature fusion. Furthermore, this study incorporates the SSFF–DySample module that utilizes DySample dynamic upsampling technology to further optimize the model ability to process features of varying scales. Finally, Slide loss is adopted as the loss function, optimizing the learning efficiency of the model for difficult-to-recognize samples, and is particularly effective in dealing with highly uneven defect distributions. Experimental validation shows that the improved YOLOv8n model enhances precision, recall, and average precision rates by 6.7%, 8%, and 8.5% respectively, with detection speeds reaching 221.7 FPS, fully meeting the requirements for high-efficiency and high-precision in solar cell panel defect detection. These improvements enhance the model performance and provide a new technical pathway for defect detection in solar cell panels.