This study focuses on enhancing impurity detection in laver images through advanced data augmentation and tailored dataset design. The irregular texture of laver and the microscopic size of impurities present significant challenges for traditional detection methods. To address these issues, a comprehensive dataset was constructed using StyleGAN to generate 10,000 synthetic laver images, coupled with traditional augmentation techniques for impurity data. The YOLOv8 model was employed for impurity detection, achieving a precision score of 0.96 under IoU 75, a substantial improvement from the 30% observed without augmented data. While occasional misclassifications of natural laver gaps as impurities were observed, the model demonstrated robust performance in detecting and excluding impurities. These results highlight the critical role of data design and augmentation in improving detection accuracy. This research provides a scalable foundation for automated impurity detection systems in the laver industry and suggests future opportunities for real-time applications and extended impurity types.