A masking apparatus according to the present invention comprises: a background image generation unit updating a background image corresponding to flow of time in an input image to generate the latest background image; a moving object extraction unit extracting a moving object from the image based on whether a change between a current image and the latest background image is equal to or greater than a predetermined threshold; a final moving object extraction unit extracting a final moving object based on whether a plurality of predetermined pixel conditions are satisfied with respect to pixels corresponding to the extracted moving object; and a face region extraction unit detecting and masking a face region from the final moving object based on distances between the central point extracted from the final moving object and a plurality of feature points and a predetermined hue saturation value (HSV) determination reference.