In this study, to address the issue of imaging distortion that occurs when visual measurements are applied in inhomogeneous refractive index environments, an inhomogeneous refractive index field produced in the inhomogeneous temperature field is developed, and the trajectory of light in the refractive index field is investigated, and the image distortion is corrected according to the trajectory's deviation. The spatial refractive index field produced using the nonuniform temperature field is reconstructed using the background schlieren method. The light emitted by the spatial point is abstracted as light rays and reproduces the light in the inhomogeneous temperature gradient environment according to the Runge-Kutta method. The world coordinates of the point are computed based on the PnP (perspective-n-point) algorithm and the pixel coordinates of the same point in the image before and after correction. Experimental findings demonstrate that the method can efficiently decrease the measurement error and correct the measured image's distortion.