A purple fringing correction method for large-area purple fringing is proposed based on color information. Considering the causes of purple fringing, the existing method for purple fringing detection is improved by making use of the chromatic aberration theory. When correcting the purple fringing, the proposed classification correction method and the method of avoiding error correction measures are proposed based on object color information, while a method for remaining purple fringing removal is corrected, so as to correct the purple fringing and reduce the error correction at the same time. Experimental results indicate that the proposed method can obtain better correction effect than the existing methods for large-area purple fringing in various scenarios of images. The corrected images are improved significantly in the subjective visual and objective data in the evaluating indicator.