Application of sobel algorithm for image edge detection in artificial mining signal analysis of fiber optic warning

With the increasing mileage of oil and gas pipeline construction, the pressure and difficulty of pipeline safety supervision are escalating. Unsafe behaviors during third-party construction near pipelines often lead to accidents. In recent years, the advancement of fiber optic early warning technology has enabled ubiquitous monitoring of third-party activities along pipelines. However, due to the complex environment along pipelines, the background noise in fiber optic vibration signals is often significant. While mechanical operations (e.g., excavators) can be detected, manual digging signals are frequently masked by noise. The Sobel operator, widely used in image edge detection for its noise resistance and suppression capabilities, was adapted to process manual digging signals. Experimental results demonstrate that the improved Sobel operator effectively suppresses noise interference, enhances the clarity of manual digging signals, and improves early warning accuracy.