An online blind pixel detection and compensation method based on scene is proposed to address the blind pixel problem of infrared focal plane detectors in infrared imaging systems. At first, 3σ criteria is used to rapidly screen suspected blind pixel targets, and then, the gradient discrimination method is used to confirm the suspected blind pixel, thereby blind pixel detection is achieved. After completing the blind pixel detection, neighborhood replacement algorithm is used to replace blind pixel and the blind pixel compensation is achieved. To verify the effectiveness of the method, simulation analysis is conducted using a constant temperature blackbody and actual scenarios. Simulation results verify that this method can achieve online blind pixel detection and compensation in infrared imaging system, and improve the quality of infrared imaging effectively.