As a key component in industrial production, industrial spools are critical for ensuring production stability and personnel safety, primarily relying on supervised learning for anomaly detection. Although supervised learning methods achieve high detection accuracy, they depend heavily on numerous manual annotations. Therefore, self-supervised learning methods emerge as potential solutions. However, traditional self-supervised methods often overly rely on the reconstruction capabilities of sub-networks when dealing with anomalous images, leading to unsatisfactory reconstruction accuracy and poor detection results. To address these issues, we propose a self-supervised anomaly detection method for industrial spool surfaces, called MN-Net. This method adapts to complex anomaly detection tasks in various industrial scenarios by using automatically generated pseudo-labels for training, eliminating the need for manual annotation. To handle interference from synthetic anomaly information caused by different feature scales, we introduce a Multi-Scale Feature Fusion (MFF) module. Additionally, to enhance the model's ability to identify anomalies, we incorporate the Neighborhood Attention (NAM) module, which significantly improves anomaly detection by focusing on local anomalies. To evaluate the detection accuracy of MN-Net, we conducted extensive experimental studies on the industrial spool dataset and the BSData dataset. The results demonstrate that MN-Net outperforms existing methods, achieving image-level AUROC (I-AUROC) scores of 96.0% and 93.8%, and pixel-based AUROC (P-AUROC) scores of 95.3% and 90.6% on the industrial spool dataset and BSData dataset, respectively.

An accurate region of interest extraction (ROI) plays an important role for both finger vein recognition systems and finger vein-based cryptography systems. In order to localize the rectangle ROI accurately, the edges of the finger and a line in the finger joint region should be detected accurately as a reference position. Because most of the existing finger edge detection methods do not work well, a robust finger edge detection method is proposed in this paper. An inner line of the finger is first detected to divide the finger vein image by two parts, after that two edge detection templates and a series of technologies such as interpolation, fit, etc. are used to detect and fix the wrong edges of the finger. Furthermore, considering that the shapes of the brighter finger joint region are irregular, multiple sliding windows including rectangle, disk, diamond and ellipse are generated, respectively to detect the reference line of the finger joint. Finally, a contour similarity distance-based method is introduced to evaluate the performance of various sliding windows. The experimental results show that the proposed edge detection method can 100% successfully detect the edges of the fingers in our finger vein image database. And for various detection windows, the ellipse window is more suitable for the detection of the finger joint reference line. So, the proposed ROI extraction method for finger vein images has a better overall performance compared with the other methods.