Abstract: Traditional manual power inspection suffers from long inspection cycles and numerous monitoring blind spots. In contrast, UAV (Unmanned Aerial Vehicle) inspection is less constrained by geographical environments and offers high inspection efficiency, making it a valuable addition to power inspection systems. As the core of UAV inspection, target tracking algorithms have become a current research hotspot. However, existing target tracking algorithms struggle to simultaneously meet the requirements for both robustness and speed, making it difficult to satisfy the demands of power inspection application scenarios. Against this backdrop, this paper proposes a multi-feature fusion-based target tracking algorithm for UAV power inspection. It utilizes HOG (Histogram of Oriented Gradients) features and color features as image inputs, trains filters through correlation filtering, and determines the tracking result based on the maximum comprehensive response. Finally, simulation experiments are conducted to verify the effectiveness and practicality of the proposed algorithm.