One of the ongoing challenges in remote sensing image processing is the destriping problem. Although numerous studies have addressed stripe noise, oblique stripe noise remains largely overlooked. Consequently, effectively removing oblique stripes from high-level remote sensing images is both a critical and unresolved task. In this work, we propose a novel destriping model capable of detecting oblique stripes regardless of their orientation, including those with low intensity. Our approach integrates the classical Hough Transform with enhanced edge detection based on a smoothing filter. A guided filter-based algorithm is then employed to restore the image information lost due to stripe noise, guided by the detected stripe patterns. Experimental results, both quantitative and qualitative, demonstrated the efficiency and robustness of the proposed method in terms of orientation estimation and stripe removal.